A method and system for adjustable resource flexibility hierarchical evaluation considering duration

Abstract

This invention relates to a method and system for hierarchical assessment of the flexibility of adjustable resources considering duration, belonging to the field of power dispatching. The invention includes: performing multi-mode coupling analysis on adjustable resources to obtain joint response probability distribution characteristics; comprehensively assessing resource regulation capabilities based on the joint response probability distribution characteristics to obtain multi-dimensional quantitative characteristics; constructing a planning constraint space based on power grid safety regulations and multi-mode physical topology constraints, and mapping the multi-dimensional quantitative characteristics to the planning constraint space to obtain a planning adaptability assessment characterization; using the planning adaptability assessment characterization as a benchmark, performing extreme value distribution pattern analysis on adjustable resources and constructing a flexibility threshold partitioning rule base; based on the flexibility threshold partitioning rule base, performing hierarchical region partitioning and confidence verification on adjustable resources to obtain optimized hierarchical labels; and generating dispatching strategies based on the optimized hierarchical labels. This invention can improve the efficiency of hierarchical assessment of the flexibility of adjustable resources.

Description

Technical Field

[0001] This invention relates to the field of power dispatching, and in particular to a method and system for hierarchical assessment of the flexibility of adjustable resources that takes into account duration. Background Technology

[0002] Existing technologies lack a complete coupled analysis mechanism for adjustable resources under multiple modes such as integrated power generation, grid, load, and storage, and direct green power connection within regional power grids. This makes it difficult to accurately extract the joint response probability distribution characteristics of resources and to perform multi-dimensional quantitative assessment of resource regulation capabilities, resulting in a lack of reliable data support for resource flexibility evaluation. Existing methods do not construct a standardized planning constraint space in conjunction with power grid safety regulations and physical topology constraints, failing to achieve precise adaptation between resource characteristics and constraints. The classification of flexibility thresholds lacks scientific basis, leading to biases and instability in the classification results. Furthermore, existing technologies do not conduct specific assessments of the duration characteristics of adjustable resources, and the classification labels lack confidence verification and optimization processes, failing to guarantee the reliability of classification conclusions. This makes it difficult to generate highly adaptable and efficient scheduling strategies based on the classification results. Therefore, improving the accuracy of adjustable resource flexibility classification assessment and the effectiveness of scheduling strategy generation has become an urgent problem to be solved. Summary of the Invention

[0003] In order to solve the above-mentioned technical problems, or at least partially solve the above-mentioned technical problems, and in view of the above-mentioned shortcomings, the present invention provides a method and system for hierarchical evaluation of adjustable resource flexibility that takes into account duration.

[0004] In a first aspect, the present invention provides a method for hierarchical evaluation of adjustable resource flexibility considering duration, comprising: The adjustable resources under multiple modes of integrated power generation, grid, load and storage and direct green power connection within the regional power grid are obtained, and multi-mode coupling analysis is performed on the adjustable resources to obtain the joint response probability distribution characteristics of the adjustable resources. Based on the joint response probability distribution characteristics, the resource adjustment capability of adjustable resources is comprehensively evaluated, and the multidimensional quantitative characteristics of adjustable resources are obtained. Based on power grid safety regulations and multi-mode physical topology constraints, a planning constraint space for adjustable resources is constructed, and multi-dimensional quantitative features are mapped to the planning constraint space to obtain a planning adaptability assessment characterization of adjustable resources. Based on the aforementioned planning adaptability assessment characterization, extreme value distribution pattern analysis is performed on adjustable resources to obtain the boundary evolution characteristics of adjustable resources. Based on the boundary evolution characteristics, a flexibility threshold division rule base for adjustable resources is constructed. Based on the flexibility threshold partitioning rule base, adjustable resources are divided into hierarchical regions to obtain preliminary hierarchical labels for adjustable resources. The confidence level of the preliminary hierarchical labels is then verified to obtain optimized hierarchical labels for adjustable resources. Based on the optimized hierarchical labels, a scheduling strategy for adjustable resources is generated.

[0005] Furthermore, the acquisition of adjustable resources under multiple modes of integrated power generation, grid-load-storage, and green electricity direct connection within the regional power grid, and the multi-mode coupling analysis of the adjustable resources to obtain the joint response probability distribution characteristics of the adjustable resources, includes: Integrate data from multiple modes of regional power grid integration, including source-grid-load-storage integration and direct green electricity connection, into adjustable resources; Time-series alignment of adjustable resources is performed to obtain standardized data for the adjustable resources; Dependency mining is performed on standardized data to obtain a multidimensional coupled state tensor of adjustable resources; Based on the multidimensional coupled state tensor, nonparametric estimation of adjustable resources is performed to obtain the dynamic response statistical characteristics of adjustable resources. The joint response probability distribution characteristics of adjustable resources are obtained by quantitatively fitting the statistical characteristics of the dynamic response.

[0006] Furthermore, based on the joint response probability distribution characteristics, a comprehensive assessment of the resource adjustment capability of adjustable resources is conducted to obtain multi-dimensional quantitative characteristics of adjustable resources, including: The joint response probability distribution characteristics are segmented using a variable-length sliding window to obtain local probability fragments of adjustable resources; Feature encoding is performed on local probability segments to obtain the temporal statistical feature spectrum of adjustable resources; Based on the time series statistical feature spectrum, the extreme value form of the adjustable indicators of adjustable resources is reconstructed to obtain the boundary description of adjustable resources. The adjustable indicators include the maximum adjustable capacity, the longest continuous response time, the average ramp rate, and the response delay variance. The boundary description is decomposed into vector orthogonal components to obtain the independent feature components of the adjustable resources; Spatial mapping of independent feature components yields multidimensional quantitative features of adjustable resources.

[0007] Furthermore, based on power grid safety regulations and multi-mode physical topology constraints, a planning constraint space for adjustable resources is constructed, and multi-dimensional quantitative features are mapped to the planning constraint space to obtain a planning adaptability assessment characterization of adjustable resources, including: The multi-mode physical topology constraints of adjustable resources are discretized and encoded to obtain the topological adjacency representation of the adjustable resources. Based on the topological adjacency representation, the power grid security regulations for adjustable resources are semantically vectorized and mapped to obtain the rule constraint representation of adjustable resources. By performing multidimensional tensor dimensionality increase on the set of rule-constrained hyperplanes, a structured weight basis for adjustable resources is obtained; By injecting topological structure into the structured weighted basis, the planning constraint space of adjustable resources is obtained; Based on the planning constraint space, the adaptability analysis of multi-dimensional quantitative characteristics is performed to obtain the planning adaptability assessment characterization of adjustable resources.

[0008] Furthermore, based on the aforementioned planning adaptability assessment characterization, extreme value distribution pattern analysis is performed on adjustable resources to obtain boundary evolution characteristics of adjustable resources. Based on these boundary evolution characteristics, a flexibility threshold partitioning rule base for adjustable resources is constructed, including: Temporal extreme value trajectories are extracted from the planning adaptability assessment representation to obtain a cluster of extreme state sequences of adjustable resources; Based on extreme state sequence clusters, the boundary of the multidimensional feature space of adjustable resources is reconstructed to obtain the boundary evolution characteristics of adjustable resources. Adaptive discretization segmentation of boundary evolution features yields an effective threshold for adjustable resources; The effective thresholds are mapped to a preset rule space to obtain a flexible threshold partitioning rule base for adjustable resources.

[0009] Furthermore, the adaptive discretization segmentation of the boundary evolution features to obtain an effective threshold for adjustable resources includes: Tensor decomposition of boundary evolution characteristics yields local mutation indices for adjustable resources; Based on the local mutation index, a topological quantization mapping is performed on the manifold surface with boundary evolution characteristics to obtain the complexity scalar field of adjustable resources. By performing abrupt change detection filtering on the complexity scalar field, the critical jump vector of adjustable resources can be obtained; By clustering the critical jump vector to convergence, the effective threshold of adjustable resources can be obtained.

[0010] Furthermore, the formula for calculating the effective threshold is as follows: ; in, For the effective threshold, As a local mutation indicator, For a scalar field of complexity, As a preset baseline threshold, This is the preset sensitivity adjustment coefficient.

[0011] Furthermore, the adjustable resources are hierarchically divided into regions based on a flexibility threshold-based rule base to obtain preliminary classification labels for the adjustable resources. The confidence level of these preliminary classification labels is then verified to obtain optimized classification labels for the adjustable resources, including: By performing time-series feature matching between the historical response records of adjustable resources and the current operating data, the continuous response duration characteristics of adjustable resources can be obtained. Based on the flexibility threshold partitioning rule base, the continuous response duration feature is hierarchically mapped to obtain the preliminary hierarchical labels of adjustable resources. Extract the resource identifiers from the preliminary classification labels, and perform correlation verification between the resource identifiers and the historical classification records of adjustable resources to obtain the label conflict records of the preliminary classification labels. Perform a hierarchical recalibration operation on the conflicting resources in the tag conflict record to obtain the recalibrated hierarchical tags of the adjustable resources; The recalibrated grading labels are compared with the multidimensional quantitative features of adjustable resources to obtain optimized grading labels for adjustable resources.

[0012] Furthermore, the step of generating a scheduling strategy for adjustable resources based on optimized hierarchical labels includes: Based on optimized hierarchical labels, the real-time operation data of adjustable resources is quantified and sorted to obtain the scheduling priority of adjustable resources. Based on scheduling priority, conflict detection is performed on the scheduling period of adjustable resources to obtain the pre-scheduling sequence of adjustable resources. Extract resource identifiers and hierarchical labels from the pre-scheduled sequence, and perform scheduling policy matching on the resource identifiers and hierarchical labels to obtain the initial scheduling policy for the adjustable resources. The consistency of the initial scheduling strategy and the joint response probability distribution characteristics is verified to obtain the verification characterization of the adjustable resources. Based on the verification representation, the initial scheduling policy is modified to obtain the scheduling policy for adjustable resources.

[0013] Secondly, the present invention provides a flexible graded assessment system for adjustable resources that takes into account duration, including: a multi-mode coupling feature analysis module, used to acquire adjustable resources under multiple modes such as integrated source-grid-load-storage and direct green power connection in the regional power grid, and to perform multi-mode coupling analysis on the adjustable resources to obtain the joint response probability distribution characteristics of the adjustable resources; The comprehensive analysis module is used to comprehensively analyze the resource adjustment capability of adjustable resources based on the joint response probability distribution characteristics, and obtain the multi-dimensional quantitative characteristics of adjustable resources. The spatial adaptation mapping module is used to construct the planning constraint space of adjustable resources based on power grid safety regulations and multi-mode physical topology constraints, and to map multi-dimensional quantitative features to the planning constraint space to obtain the planning adaptability assessment characterization of adjustable resources. The rule construction module is used to analyze the extreme value distribution pattern of adjustable resources based on the planning adaptability assessment characterization, obtain the boundary evolution characteristics of adjustable resources, and construct a rule base for dividing the flexibility threshold of adjustable resources based on the boundary evolution characteristics. The hierarchical verification and optimization module is used to divide the adjustable resources into hierarchical regions based on the flexibility threshold rule base, obtain the preliminary hierarchical labels of the adjustable resources, and perform confidence verification on the preliminary hierarchical labels to obtain the optimized hierarchical labels of the adjustable resources. The strategy generation module is used to generate scheduling strategies for adjustable resources based on optimization level labels.

[0014] The technical solutions provided in the embodiments of the present invention have the following advantages compared with the prior art: This invention, through multi-mode coupling analysis and multi-dimensional quantitative judgment, fully extracts the joint response probability distribution characteristics of adjustable resources, accurately characterizes the multi-dimensional quantitative features of resource adjustment capabilities, and improves the completeness and accuracy of adjustable resource assessment data. Based on the construction of planning constraint space and boundary evolution feature analysis, a scientific rule base for flexibility threshold classification is established to ensure the rationality of resource classification criteria and the stability of the classification process.

[0015] This invention outputs reliable resource classification results through a graded label confidence verification and optimization process. Based on the optimized graded labels, a scheduling strategy is generated to improve the adaptability and execution efficiency of the scheduling strategy, ensuring the coordinated stability and efficient controllability of the integrated operation of power generation, grid, load and storage, and direct connection of green electricity in the regional power grid. Attached Figure Description

[0016] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.

[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 A flowchart of a method for hierarchical evaluation of adjustable resource flexibility considering duration, provided in an embodiment of the present invention; Figure 2 A flowchart illustrating the joint response probability distribution characteristics of adjustable resources, provided in an embodiment of the present invention; Figure 3 A flowchart for obtaining multidimensional quantization features of adjustable resources provided in an embodiment of the present invention; Figure 4 A flowchart for obtaining a planning adaptability assessment characterization of adjustable resources, provided as an embodiment of the present invention; Figure 5 A flowchart for constructing a flexibility threshold partitioning rule base for adjustable resources provided in an embodiment of the present invention; Figure 6 A flowchart for obtaining optimized hierarchical labels for adjustable resources provided in an embodiment of the present invention; Figure 7 A flowchart illustrating the generation of a scheduling strategy for adjustable resources based on optimized hierarchical labels, provided in an embodiment of the present invention. Figure 8 A threshold distribution curve of adjustable resource joint response flexibility considering duration is provided in an embodiment of the present invention. Figure 9 This invention provides a threshold distribution curve for the flexibility of adjustable resource boundary evolution, taking into account duration. Figure 10 This is a schematic diagram of the structure of an adjustable resource flexibility classification assessment system that takes into account the duration of the event, according to an embodiment of the present invention. Figure 11 This is a schematic diagram of an adjustable resource flexibility grading assessment device that takes into account duration, provided in an embodiment of the present invention. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0020] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0021] Example 1 like Figure 1 As shown, the present invention provides a method for hierarchical evaluation of adjustable resource flexibility considering duration, comprising: This study acquires adjustable resources under multiple modes of integrated power generation, grid-load-storage, and green energy direct connection within the regional power grid, and performs multi-mode coupling analysis on these adjustable resources to obtain the joint response probability distribution characteristics. For example... Figure 2 As shown, the specific process includes: The data from the regional power grid under the integrated source-grid-load-storage and green electricity direct connection modes will be integrated into adjustable resources. Specifically, all operational data collected under the integrated source-grid-load-storage and green electricity direct connection modes in the regional power grid will be uniformly collected and organized. The data will be classified and merged according to resource type and operation scenario, and operational information from different sources and of different types will be fully integrated to eliminate format differences and source barriers between data, forming a comprehensive, complete and unified adjustable resource.

[0022] Adjustable resources are time-series aligned to obtain standardized data. Specifically, the integrated adjustable resources are matched and arranged according to a unified timeline. The time nodes corresponding to each data item are checked one by one. Discontinuous time parts are filled in, and misaligned time parts are corrected and put back in place, so that all resource data are synchronized and consistent in the time dimension, forming standardized data of adjustable resources that are standardized, unified and complete.

[0023] Dependency mining is performed on standardized data to obtain a multidimensional coupling state tensor of adjustable resources. Specifically, the relationships between each set of processed standardized data are identified, and the interaction and synergistic change patterns between various data are comprehensively analyzed. The mutual influence relationships and linkage states between different resources are extracted, and the mutual influence relationships and linkage states are arranged in an orderly manner according to a multidimensional structure to construct a multidimensional coupling state tensor of adjustable resources that can fully reflect the resource coupling state.

[0024] Based on the multidimensional coupled state tensor, nonparametric estimation of adjustable resources is performed to obtain the dynamic response statistical characteristics of adjustable resources. Specifically, based on the multidimensional coupled state tensor, data statistics are carried out according to the actual distribution of resource response. The statistical calculation of response characteristics is completed based on the actual distribution of the sample data itself, comprehensively recording the response performance of resources under different operating conditions, and obtaining the dynamic response statistical characteristics of adjustable resources that fully reflect the resource response law.

[0025] The dynamic response statistical characteristics are quantitatively fitted to obtain the joint response probability distribution characteristics of adjustable resources. The quantitative fitting performs numerical normalization and trend restoration on the obtained dynamic response statistical characteristics, gradually transforming the discrete distribution statistical results into a continuous and smooth distribution form. This fully restores the response probability law of resources under multi-mode operation scenarios, ensuring that the distribution results closely match the actual operating state. Finally, it forms an accurate, complete, and true reflection of the resource response characteristics, resulting in the joint response probability distribution characteristics of adjustable resources.

[0026] By systematically integrating data, time-series alignment, dependency mining, non-parametric estimation, and quantization fitting processes, we can stably and efficiently obtain accurate and reliable joint response probability distribution characteristics of adjustable resources. This provides solid data support for subsequent assessment of resource adjustment capabilities and improves the stability and accuracy of the entire evaluation process.

[0027] Based on the joint response probability distribution characteristics, a comprehensive assessment of the resource adjustment capability of adjustable resources is conducted, resulting in multi-dimensional quantitative characteristics of adjustable resources. For example... Figure 3 As shown, the specific process includes: The joint response probability distribution characteristics are segmented by a variable-length sliding window to obtain local probability fragments of adjustable resources. Specifically, window intervals of different lengths are set according to the actual changing trend of the joint response probability distribution characteristics and the data density. The data segments are truncated segment by segment along the data sequence with dynamically changing window lengths. Each segment of data retains the probability distribution information within the corresponding interval, forming local probability fragments of adjustable resources that can accurately reflect the local distribution state.

[0028] Feature encoding is performed on local probability segments to obtain the temporal statistical feature spectrum of adjustable resources. Specifically, structured numerical transformation and feature extraction are performed on each local probability segment, and information such as the probability change trend, numerical fluctuation amplitude, distribution concentration, etc. within the segment are transformed into an ordered feature sequence. The feature sequences corresponding to all local probability segments are combined and arranged in chronological order to form a complete temporal statistical feature spectrum of adjustable resources that presents the temporal change pattern.

[0029] Based on time-series statistical feature spectra, the extreme value forms of adjustable indicators of adjustable resources are reconstructed to obtain the boundary description of adjustable resources. The adjustable indicators include maximum adjustable capacity, longest continuous response time, average ramp rate, and response delay variance. Specifically, based on the time-series statistical feature spectra, numerical restoration and form construction are performed on four adjustable indicators: maximum adjustable capacity, longest continuous response time, average ramp rate, and response delay variance. The extreme values ​​and change boundaries of each indicator during operation are determined one by one. The extreme states and boundary information of all indicators are integrated to form a complete boundary description of adjustable resources that defines the adjustment range of resources.

[0030] The boundary description is decomposed into vector orthogonal components to obtain independent feature components of the adjustable resource. Specifically, all feature information contained in the boundary description is split into mutually independent and non-interfering directions, and the mixed composite features are decomposed into independent feature units of a single dimension, so that each unit corresponds to only one independent resource attribute, thus obtaining independent feature components of the adjustable resource with clear attributes that do not affect each other.

[0031] Spatial mapping of independent feature components yields multidimensional quantitative features of adjustable resources. Specifically, all independent feature components are positioned and projected according to a preset multidimensional coordinate system, transforming the numerical attributes of each component into positional information in multidimensional space. This ensures that each feature forms a comparable and quantifiable expression within a unified space, ultimately resulting in multidimensional quantitative features of adjustable resources that comprehensively, completely, and accurately characterize resource adjustment capabilities.

[0032] By using a variable-length sliding window to segment feature encoding, extreme value morphology reconstruction vector orthogonal decomposition, and spatial mapping, a coherent processing flow can be established. This can fully quantify the adjustment capability of adjustable resources, clearly define the resource operation boundary, and provide a stable and reliable quantitative basis for subsequent planning constraint matching and flexibility classification.

[0033] Based on power grid safety regulations and multi-mode physical topology constraints, a planning constraint space for adjustable resources is constructed, and multi-dimensional quantitative features are mapped to the planning constraint space to obtain a planning adaptability evaluation characterization of adjustable resources; in the embodiments of this invention, such as Figure 4 As shown, the specific process includes: The multi-mode physical topology constraints of adjustable resources are discretized and encoded to obtain the topological adjacency representation of adjustable resources. Specifically, the physical topology constraints of adjustable resources under various operating modes such as source-grid-load-storage integration and direct green power connection are decomposed one by one. The continuous physical connection relationship, equipment carrying boundary, line transmission constraint, and node connectivity constraint are all transformed into independent and identifiable discrete units. These discrete units are then arranged and combined in an orderly manner according to the order of actual physical connection and adjacency relationship, and the association information and constraint boundary of all topology constraints are completely preserved to form a topological adjacency representation of adjustable resources that can accurately describe the topological association and constraint state of resources.

[0034] Based on topological adjacency representation, semantic vectorization mapping is performed on the power grid safety regulations for adjustable resources to obtain the rule constraint representation of adjustable resources. Specifically, based on the topological adjacency representation framework, all rule contents included in the power grid safety regulations, such as operating limits, safety boundaries, operating procedures, fault constraints, and stability requirements, are structurally transformed one by one. Each text-based safety rule is converted into a fixed-dimensional numerical vector, so that the constraint strength, scope of application, and limiting conditions of the rule are all clearly presented in vector form. All rule vectors are combined to form a complete rule constraint representation of adjustable resources that carries all the constraint information of power grid safety.

[0035] The set of rule-constrained hyperplanes is subjected to multidimensional tensor dimensionality upscaling to obtain a structured weight basis for adjustable resources. Specifically, the multiple rule-constrained hyperplanes formed by the rule-constraint representations are expanded in dimension and constructed hierarchically according to a higher-dimensional feature space. The original low-dimensional single constraint plane is gradually upgraded to a multidimensional tensor structure. During the dimensionality upscaling process, the correlation and priority between each constraint are fully preserved, so that all constraints form a clear hierarchical and closely related overall structure, resulting in a structured weight basis for adjustable resources with a stable structure and clear weight distribution.

[0036] By injecting topology structure into the structured weight base, a planning constraint space for adjustable resources is obtained. Specifically, all physical topology connections, node levels, line constraints, and equipment location information contained in the topology adjacency representation are fully injected into the structured weight base, so that each weight unit of the structured weight base corresponds precisely to the actual physical topology nodes and connections. This allows the constraints to be deeply integrated with the physical topology restrictions, forming a planning constraint space for adjustable resources that covers all restrictions of power grid safety regulations and multi-mode physical topology.

[0037] Based on the planning constraint space, adaptability analysis is performed on multi-dimensional quantitative features to obtain a planning adaptability assessment characterization of adjustable resources. Specifically, using the constructed planning constraint space as a unified evaluation benchmark, the multi-dimensional quantitative features of the previously obtained adjustable resources are placed into this space one by one for adaptation and verification. The degree of fit between each quantitative feature and the constraint boundary, the compatibility with topological constraints, and the compliance with safety rules are analyzed. All adaptation results are combined to form a planning adaptability assessment characterization of adjustable resources that can comprehensively reflect the adaptability level of resources under the planning scenario.

[0038] Through a complete process of discretization encoding, semantic vectorization mapping, multidimensional tensor dimensionality increase, topology injection and adaptability analysis, a planning constraint space that fits the actual operating conditions of the power grid can be accurately constructed, and the adaptation judgment of multidimensional quantitative features and constraints can be efficiently completed, providing a rigorous and reliable evaluation basis for subsequent flexibility threshold division.

[0039] Based on the planning adaptability assessment, extreme value distribution patterns of adjustable resources are analyzed to obtain their boundary evolution characteristics. Based on these boundary evolution characteristics, a rule base for classifying the flexibility thresholds of adjustable resources is constructed. Figure 5 As shown, the specific process includes: The extreme value trajectory of the planning adaptability assessment is extracted to obtain the extreme state sequence cluster of adjustable resources. Specifically, based on all the data of the planning adaptability assessment, the entire time series data sequence is traversed at a fixed time step. In each time step, the maximum and minimum values ​​of each dimension feature at the current moment are searched and determined. Each set of extreme values ​​is recorded and connected in chronological order to form a continuous change trajectory. The extreme value trajectories of different resource types and different operating modes are uniformly collected and organized to form the extreme state sequence cluster of adjustable resources that includes the extreme states of all dimensions and all time periods.

[0040] Based on extreme state sequence clusters, the boundary of the multidimensional feature space of adjustable resources is reconstructed to obtain the boundary evolution characteristics of adjustable resources. Specifically, based on all extreme point data in the extreme state sequence clusters, the feasible domain of the resources is redefined within the pre-constructed multidimensional feature space. The extreme points in each dimension are used as boundary control points. The discrete extreme points are connected into a smooth and continuous closed boundary through spatial fitting. At the same time, the position offset and shape change of the boundary under different operating conditions and time periods are recorded in conjunction with temporal changes, so as to fully present the dynamic evolution process of the resource boundary and form the boundary evolution characteristics of adjustable resources that can accurately describe the dynamic change law of the extreme state of resources.

[0041] An adaptive discretization segmentation method is used to segment boundary evolution features to obtain effective thresholds for adjustable resources. The specific process includes: tensor decomposition of the boundary evolution features to obtain local mutation indices for adjustable resources; specifically, multi-dimensional and multi-level tensor decomposition of the overall boundary evolution features is performed, decomposing the high-dimensional overall boundary features into multiple low-dimensional local feature components. Within each local feature component, the location with the most drastic numerical change is identified, and the magnitude and rate of change at that location are calculated. This value visually represents the degree of feature mutation in the local region, ultimately forming a local mutation index for adjustable resources that accurately reflects the intensity of local boundary changes. Based on the local mutation index, topological quantization mapping is performed on the manifold surface of the boundary evolution features to obtain a complexity scalar field for adjustable resources. Specifically, based on the obtained local mutation index, the manifold surface corresponding to the boundary evolution features is quantized point-by-point according to spatial location. The local mutation values, topological connections, and direction of change at each spatial location are uniformly converted into scalar values. All scalar values ​​are then arranged according to spatial distribution rules to form a continuous field structure, fully presenting the complex distribution of the overall boundary changes. Furthermore, a complexity scalar field capable of comprehensively characterizing the complexity of adjustable resources with varying boundary conditions is obtained. Abrupt change point detection filtering is applied to the complexity scalar field to obtain the critical jump vector of the adjustable resources. Specifically, all scalar data within the complexity scalar field are scanned and detected point-by-point to accurately identify key points where significant numerical jumps occur. Through filtering, interference and outliers caused by data fluctuations are eliminated, retaining only valid points that truly reflect changes in the critical boundary state. These valid points are then combined according to their corresponding spatial dimensions and temporal order to form directional feature vectors, resulting in clearly labeled vectors. The critical transition vectors of adjustable resources are determined based on their boundary critical transition positions and intensities. Clustering convergence is performed on these critical transition vectors to obtain an effective threshold for the adjustable resources. Specifically, all critical transition vectors are compared for feature similarity, and vectors with similar values ​​and consistent trends are grouped into the same cluster. For each cluster, center calculation and iterative convergence are performed on the vectors within that cluster, ensuring that each cluster consistently converges to a unique center value. This center value objectively represents the standard threshold for the corresponding critical state, ultimately yielding a stable, reliable, and practically instructive effective threshold for the adjustable resources. The formula for calculating the effective threshold is as follows: ; Among them, among them, For the effective threshold, As a local mutation indicator, For a scalar field of complexity, As a preset baseline threshold, This is a preset sensitivity adjustment coefficient. The effective threshold is used to characterize the effective judgment boundary of the signal within the current analysis area. By combining the basic threshold benchmark, local mutation index, complexity scalar field, and sensitivity adjustment coefficient, it comprehensively reflects the fluctuation characteristics and distribution complexity of the signal within the area, providing a unified judgment standard for subsequent signal validity determination. The basic threshold benchmark is obtained from the statistical analysis of historical system operation data. By extracting threshold samples under long-term stable operation, removing extreme conditions and abnormal samples, the statistical mean of all effective samples is taken as the basic threshold benchmark. The local mutation index is obtained by comparing the signal fluctuation amplitude point by point within the current analysis area. It iterates through all data points within the area, calculates the numerical difference between adjacent data points, and takes the maximum value among all differences as the local mutation index. The complexity scalar field is obtained by quantifying the dispersion of the data distribution within the current analysis area. It iterates through all data points within the area, calculates the sum of numerical differences between each data point and its neighboring data points, and takes the average of all sums of differences as the complexity scalar field. The sensitivity adjustment coefficient is determined by the system's preset adjustment rules. Based on the detection requirements of the actual application scenario, a fixed value is selected within a preset range to adjust the influence of local mutation indicators on the effective threshold. When the local mutation indicator increases, the effective threshold increases accordingly. When the complexity scalar field increases, the effective threshold decreases accordingly. When the baseline threshold increases, the effective threshold increases accordingly. When the sensitivity adjustment coefficient increases, the effective threshold increases accordingly. The influence direction of each parameter on the effective threshold remains stable, ensuring that the effective threshold accurately reflects the actual change pattern of the signal within the region.

[0042] Valid thresholds are mapped to a predefined rule space to obtain a flexible threshold classification rule library for adjustable resources. Specifically, each valid threshold is mapped to a predefined rule space. According to the hierarchical framework, applicable conditions, constraint logic, and calling format of the rule space, corresponding hierarchical judgment conditions, applicable scenario scope, and execution logic are configured for each valid threshold. All thresholds and corresponding rules are structured, integrated, and uniformly stored to form a complete and standardized flexible threshold classification rule library for adjustable resources that can be directly used for hierarchical judgment.

[0043] Through a full-process processing including temporal extreme value trajectory extraction, multi-dimensional feature space boundary reconstruction, tensor decomposition, topological quantization mapping, mutation point detection filtering, cluster center convergence and regularized spatial mapping, the boundary evolution features of adjustable resources can be accurately extracted, effective thresholds can be stably obtained, and a scientific and rigorous flexible threshold division rule base can be constructed. This provides reliable rule support and judgment basis for subsequent resource classification, improving the stability and accuracy of the classification process.

[0044] Figure 8This section compares the actual measured values ​​(solid blue line) and theoretical values ​​(dashed black line) of the effective threshold for different resource IDs. The curves show significant fluctuations; the effective thresholds for some resource IDs (such as 4 and 9) are significantly lower than the theoretical values, while those for others (such as 5 and 8) are higher. Overall, the curves exhibit irregular fluctuations, reflecting the substantial differences in the effective thresholds of various resource units in the actual system. These differences may be influenced by factors such as load, efficiency, or external interference, leading to deviations between actual performance and theoretical expectations.

[0045] Figure 9 The relationship between resource number and effective threshold is also presented, but the fluctuation trend is more extreme. The first half (resources 1-4) is relatively stable, close to the theoretical value; the middle half (5-7) decreases slightly; the second half (8-11) rises sharply, with resource 11 reaching a peak, far exceeding the theoretical value, while resource 12 drops sharply to the lowest point. This drastic fluctuation may indicate that the system is performing abnormally efficiently or overloaded on high-load resources (such as 8 and 11), while resource 12 may have a fault or data anomaly. Overall, it reflects that the system's performance is extremely unstable across different resources, and the causes of the high values ​​and sharp drops need to be closely monitored.

[0046] Based on a rule base for classifying flexible thresholds, adjustable resources are divided into hierarchical regions to obtain preliminary classification labels. These preliminary labels are then validated with confidence to obtain optimized classification labels for the adjustable resources. Figure 6 As shown, the specific process includes: The historical response records of adjustable resources are matched with the current operating data to obtain the continuous response duration characteristics of adjustable resources. Specifically, the historical response records of adjustable resources and the current operating data are aligned segment by segment along the same time dimension. The historical response status and the current operating status at the same time node are compared one by one to extract key time-series characteristics such as the duration, continuous stability, and time period overlap during the continuous response process. By matching point by point throughout the entire time period, the complete time interval and continuous maintenance status of the resource from the start of response to the end of response are determined, forming the continuous response duration characteristics of adjustable resources that accurately reflect the actual response duration capability of the resources.

[0047] Based on the flexibility threshold classification rule base, the continuous response duration feature is classified and mapped to obtain the preliminary classification label of adjustable resources. Specifically, according to the established threshold standards and classification logic in the flexibility threshold classification rule base, the value of the continuous response duration feature is compared with the thresholds of each level in the rule base item by item. The threshold range in which the feature value is located is directly mapped to the preset classification result, and each adjustable resource is assigned a unique initial classification identifier, forming a preliminary classification label of adjustable resources that intuitively reflects the preliminary classification result of the resources.

[0048] Extract resource identifiers from the preliminary classification labels and verify the association between these resource identifiers and the historical classification records of adjustable resources to obtain label conflict records for the preliminary classification labels. Specifically, extract a unique resource identity identifier from each preliminary classification label and verify the association between this resource identifier and all previous historical classification records of that resource. Compare the current preliminary classification results with the historical classification results to see if there are any inconsistencies in the levels, conflicts in the judgment conditions, or contradictions in the classification results. Record all inconsistencies and contradictions completely to form a label conflict record for adjustable resources that accurately reflects the classification differences.

[0049] A hierarchical recalibration operation is performed on the conflicting resources in the label conflict record to obtain the recalibrated hierarchical labels of the adjustable resources. Specifically, for the adjustable resources marked with hierarchical conflicts in the label conflict record, all core data such as continuous response duration characteristics, multi-dimensional quantitative characteristics, and planning adaptability assessment characteristics are retrieved again. The complete hierarchical judgment process is re-executed according to the flexibility threshold rule base, the judgment bias of the conflict part is corrected, and the hierarchical result that conforms to all data and rules is re-determined to obtain the recalibrated hierarchical labels of the adjustable resources that have eliminated conflicts and have sufficient judgment basis.

[0050] The recalibrated grading labels are compared with the multidimensional quantitative features of adjustable resources to obtain optimized grading labels for the adjustable resources. Specifically, the recalibrated grading labels are compared with the previously generated multidimensional quantitative features of adjustable resources item by item to verify whether the grading results match the values ​​of quantitative features such as maximum adjustable capacity, longest sustained response time, average ramp rate, and response delay variance. This confirms that the grading level is consistent with the actual resource capabilities, eliminates all mismatches, and finally forms optimized grading labels for adjustable resources that are stable, reliable, and have high confidence.

[0051] Through a complete process of time-series feature matching, hierarchical mapping, association verification, hierarchical recalibration and consistency comparison, the continuous response duration features can be accurately obtained and the initial hierarchical classification can be completed. Hierarchical conflicts can be effectively identified and corrected, significantly improving the accuracy and reliability of hierarchical labels, and providing a solid and reliable hierarchical foundation for the generation of subsequent scheduling strategies.

[0052] Based on the optimized hierarchical labels, a scheduling strategy for adjustable resources is generated. For example... Figure 7 As shown, the specific process includes: Based on optimized hierarchical labels, the real-time operation data of adjustable resources is quantitatively sorted to obtain the scheduling priority of adjustable resources. Specifically, based on the resource level and capability attributes determined by the optimized hierarchical labels, the real-time operation data of adjustable resources are numerically calculated and sorted item by item. They are arranged from high to low according to core indicators such as resource adjustment capability, response speed, and duration, clarifying the order and importance of each resource in scheduling, and obtaining a clear and standardized scheduling priority of adjustable resources.

[0053] Based on scheduling priority, conflict detection is performed on the scheduling time periods of adjustable resources to obtain a pre-scheduling sequence of adjustable resources. Specifically, based on the order defined by scheduling priority, the time intervals of each adjustable resource plan to participate in scheduling are compared and overlap identified one by one. It is detected whether there are scheduling time conflicts, capacity conflicts, and topology constraint conflicts among different resources in the same time period. All conflict locations and conflict types are marked. After eliminating conflicts, the resources are arranged in order of priority to obtain a pre-scheduling sequence of adjustable resources with reasonable timing and no conflicts.

[0054] Resource identifiers and hierarchical tags are extracted from the pre-scheduling sequence, and scheduling strategies are matched between the resource identifiers and hierarchical tags to obtain the initial scheduling strategy for adjustable resources. Specifically, the resource identifier and corresponding optimized hierarchical tag for each adjustable resource are extracted one by one from the pre-scheduling sequence. The resource type, access location, and operation restrictions are determined based on the resource identifier. The scheduling intensity, duration, and adjustment range are determined in combination with the hierarchical tag. The matching is completed one by one according to the grid scheduling requirements and resource adaptation conditions to form the initial scheduling strategy for adjustable resources that meets the priority and operating condition requirements.

[0055] The consistency of the initial scheduling strategy with the joint response probability distribution characteristics is verified to obtain the verification characterization of adjustable resources. Specifically, the scheduling period, scheduling capacity, scheduling rhythm and other contents in the initial scheduling strategy are checked item by item with the previously generated joint response probability distribution characteristics to verify whether the scheduling arrangement conforms to the response rules, probability distribution and coordination characteristics of resource multi-mode. The matching degree and deviation are recorded to form a verification characterization of adjustable resources that objectively reflects the degree of strategy adaptability.

[0056] Based on the verification representation, the initial scheduling strategy is revised to obtain a scheduling strategy for adjustable resources. Specifically, based on the matching results and deviation information presented by the verification representation, the contents of the initial scheduling strategy that do not conform to the response pattern or have adaptation deviations are adjusted and corrected. The scheduling period, scheduling capacity, and coordination mode are redefined to eliminate the contradiction between the strategy and resource characteristics, and finally a safe, reliable, and efficient scheduling strategy for adjustable resources is formed.

[0057] Through a complete process of quantitative sorting, conflict detection, strategy matching, consistency verification and strategy correction, it is possible to generate scheduling strategies that fit resource characteristics and meet power grid operation requirements based on optimized hierarchical labels, thereby improving the rationality, feasibility and execution efficiency of scheduling arrangements and ensuring the stability and controllability of the power grid under multi-mode operation.

[0058] Example 2 like Figure 10 As shown, an embodiment of the invention provides a hierarchical evaluation system for adjustable resource flexibility considering duration. This hierarchical evaluation system 100 can be installed in an electronic device. Depending on the functions implemented, the hierarchical evaluation system 100 may include a multi-mode coupling feature parsing module 101, a comprehensive judgment module 102, a spatial adaptation mapping module 103, a rule construction module 104, a hierarchical verification and optimization module 105, and a strategy generation module 106. The modules described in this invention can also be called units, referring to a series of computer program segments that can be executed by the processor of an electronic device and perform a fixed function, stored in the memory of the electronic device.

[0059] In this embodiment, the functions of each module / unit are as follows: The multi-mode coupling feature analysis module 101 is used to obtain adjustable resources under multiple modes of integrated source-grid-load-storage and direct green power connection in the regional power grid, and to perform multi-mode coupling analysis on the adjustable resources to obtain the joint response probability distribution characteristics of the adjustable resources. The comprehensive judgment module 102 is used to comprehensively judge the resource adjustment capability of adjustable resources based on the joint response probability distribution characteristics, and obtain the multi-dimensional quantitative characteristics of adjustable resources. The spatial adaptation mapping module 103 is used to construct a planning constraint space for adjustable resources based on power grid safety regulations and multi-mode physical topology constraints, and to map multi-dimensional quantitative features to the planning constraint space to obtain a planning adaptability evaluation characterization of adjustable resources. The rule construction module 104 is used to perform extreme value distribution pattern analysis on adjustable resources based on the planning adaptability assessment characterization, obtain the boundary evolution characteristics of adjustable resources, and construct a flexible threshold division rule base for adjustable resources based on the boundary evolution characteristics. The hierarchical verification and optimization module 105 is used to divide the adjustable resources into hierarchical regions based on the flexibility threshold division rule base, obtain the preliminary hierarchical labels of the adjustable resources, and perform confidence verification on the preliminary hierarchical labels to obtain the optimized hierarchical labels of the adjustable resources. The strategy generation module 106 is used to generate a scheduling strategy for adjustable resources based on the optimization hierarchical labels.

[0060] Example 3 like Figure 11 As shown, this embodiment of the invention provides an adjustable resource flexibility grading assessment device considering duration, comprising: at least one processing unit, the processing unit being connected to a storage unit via a bus unit, the storage unit serving as a computer-readable storage medium, and used to store software programs, computer-executable programs, and modules, such as the software program, computer-executable program, and module corresponding to the adjustable resource flexibility grading assessment method considering duration in this embodiment of the invention. The processing unit implements the aforementioned adjustable resource flexibility grading assessment method considering duration by running the software program, computer-executable program, and module stored in the storage unit.

[0061] Of course, the computer program stored in the storage unit of the adjustable resource flexibility classification assessment device considering duration provided in the embodiments of the present invention is not limited to the method operation described above, but can also execute related operations in the adjustable resource flexibility classification assessment method considering duration provided in any embodiment of the present invention.

[0062] The execution entity of the adjustable resource flexibility grading assessment method considering duration includes, but is not limited to, at least one of the following electronic devices that can be configured to execute the method provided in the embodiments of this application: a server, a terminal, etc. In other words, the adjustable resource flexibility grading assessment method considering duration can be executed by software or hardware installed on a terminal device or a server device. The server includes, but is not limited to, a single server, a server cluster, a cloud server, or a cloud server cluster. The server can be an independent server or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks (CDN), and big data and artificial intelligence platforms.

[0063] In the embodiments provided by this invention, it should be understood that the disclosed structures and methods can be implemented in other ways. For example, the structural embodiments described above are merely illustrative. For instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection through some interfaces, structures, or units, and may be electrical, mechanical, or other forms.

[0064] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0065] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0066] The above description is merely a specific embodiment of the present invention, enabling those skilled in the art to understand or implement the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the present invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features claimed herein.

Claims

1. A method for hierarchical assessment of the flexibility of adjustable resources considering duration, characterized in that, include: This study acquires adjustable resources under multiple modes of integrated power generation, grid-load, and energy storage, as well as direct green energy connection within the regional power grid. It then performs multi-mode coupling analysis on these adjustable resources to obtain their joint response probability distribution characteristics. This includes: integrating data from these modes into adjustable resources; performing time-series alignment to obtain standardized data; mining dependencies in the standardized data to obtain a multidimensional coupling state tensor for the adjustable resources; identifying relationships among groups of standardized data to comprehensively analyze the interactions and synergistic changes between different data points; extracting mutual influence relationships and linkage states between different resources; and constructing a multidimensional coupling state tensor that fully reflects the resource coupling state. Based on this multidimensional coupling state tensor, non-parametric estimation of the adjustable resources is performed to obtain their dynamic response statistical characteristics. Finally, quantitative fitting of these dynamic response statistical characteristics yields the joint response probability distribution characteristics of the adjustable resources. Based on the joint response probability distribution characteristics, the resource adjustment capability of adjustable resources is comprehensively assessed, resulting in multi-dimensional quantitative characteristics of adjustable resources. These characteristics include: segmenting the joint response probability distribution characteristics using a variable-length sliding window to obtain local probability segments of the adjustable resources; encoding the local probability segments to obtain the temporal statistical feature spectrum of the adjustable resources; reconstructing the extreme value form of the adjustable indicators based on the temporal statistical feature spectrum to obtain the boundary description of the adjustable resources, including maximum adjustable capacity, longest sustained response time, average ramp rate, and response delay variance; performing vector orthogonal decomposition on the boundary description to obtain independent feature components of the adjustable resources; and spatial mapping of the independent feature components to obtain the multi-dimensional quantitative characteristics of the adjustable resources. Based on power grid safety regulations and multi-mode physical topology constraints, a planning constraint space for adjustable resources is constructed, and multi-dimensional quantitative features are mapped to the planning constraint space to obtain a planning adaptability evaluation representation of adjustable resources. This includes: discretizing and encoding the multi-mode physical topology constraints of adjustable resources to obtain a topological adjacency representation of the adjustable resources; performing semantic vectorization mapping on the power grid safety regulations of adjustable resources based on the topological adjacency representation to obtain a rule constraint representation of the adjustable resources; performing multi-dimensional tensor dimensionality increase on the rule constraint hyperplane group to obtain a structured weighted basis for the adjustable resources; injecting topological structure into the structured weighted basis to obtain a planning constraint space for adjustable resources; and performing adaptability analysis on the multi-dimensional quantitative features based on the planning constraint space to obtain a planning adaptability evaluation representation of the adjustable resources. Based on the planning adaptability assessment characterization, extreme value distribution morphology analysis is performed on adjustable resources to obtain boundary evolution characteristics. Based on these boundary evolution characteristics, a flexibility threshold classification rule base for adjustable resources is constructed, including: extracting time-series extreme value trajectories from the planning adaptability assessment characterization to obtain extreme state sequence clusters of adjustable resources; and reconstructing the boundary of the multidimensional feature space of adjustable resources based on these extreme state sequence clusters to obtain boundary evolution characteristics. Specifically, based on all extreme value point data in the extreme state sequence clusters, the feasible domain of the resources is redefined within the pre-constructed multidimensional feature space, and the extreme value points in each dimension are used as boundary control points. Discrete extreme values ​​are then combined using spatial fitting. The value points are connected to form a smooth and continuous closed boundary. Simultaneously, the temporal variation records the boundary's positional shift and morphological changes under different operating conditions and time periods, fully presenting the dynamic evolution process of the resource boundary. This forms the boundary evolution characteristics of adjustable resources that can accurately describe the dynamic changes in the resource's extreme state. Adaptive discretization segmentation of the boundary evolution characteristics yields effective thresholds for adjustable resources. This process includes: tensor decomposition of the boundary evolution characteristics to obtain local mutation indices for adjustable resources; specifically, multi-dimensional and multi-level tensor decomposition of the overall boundary evolution characteristics, decomposing the high-dimensional overall boundary features into multiple low-dimensional local features. The system identifies the most drastic numerical changes within each local feature component, calculates the magnitude and rate of change at that location, and visually reflects the degree of feature mutation in the local region. This results in a local mutation index for adjustable resources that accurately reflects the intensity of local boundary changes. Based on this index, the manifold surface of the boundary evolution features is topologically quantized and mapped to obtain a scalar field of complexity for the adjustable resources. Specifically, based on the obtained local mutation index, the manifold surface corresponding to the boundary evolution features is quantized point-by-point according to spatial location. The local mutation values, topological connectivity, and direction of change at each spatial location are uniformly converted into scalar values. Finally, all scalar values ​​are distributed according to spatial distribution rules. The law arrangement forms a continuous field structure, fully presenting the complex distribution of the boundary as a whole, resulting in a complexity scalar field of adjustable resources that can comprehensively characterize the complexity of boundary changes. Abrupt change point detection filtering is applied to the complexity scalar field to obtain the critical jump vector of the adjustable resources. Specifically, all scalar data within the complexity scalar field are scanned and detected point by point to identify key points where significant numerical jumps occur. Through filtering, interference and outlier points caused by data fluctuations are removed, retaining effective points that truly reflect the changes in the critical boundary state. These effective points are combined according to their corresponding spatial dimensions and temporal order into directional feature vectors, resulting in the critical jump vector of the adjustable resources that identifies the location and intensity of the critical boundary jump.The effective threshold for adjustable resources is obtained by clustering and convergence of critical transition vectors. Specifically, feature similarity is compared among all critical transition vectors, and vectors with similar values ​​and consistent trends are grouped into the same cluster. For each cluster, center calculation and iterative convergence are performed to ensure that each cluster converges stably to a unique center value, representing the standard threshold for the corresponding critical state. This yields a stable, reliable, and practically instructive effective threshold for adjustable resources. The effective threshold is then mapped to a predefined rule space to obtain a rule base for classifying the flexibility threshold of adjustable resources. Based on the flexibility threshold partitioning rule base, adjustable resources are divided into hierarchical regions to obtain preliminary hierarchical labels for adjustable resources. The confidence level of the preliminary hierarchical labels is then verified to obtain optimized hierarchical labels for adjustable resources. Based on the optimized hierarchical labels, a scheduling strategy for adjustable resources is generated.

2. The adjustable resource flexibility grading assessment method considering duration as described in claim 1, characterized in that, The formula for calculating the effective threshold is as follows: ； in, For the effective threshold, As a local mutation indicator, For a scalar field of complexity, As a preset baseline threshold, This is the preset sensitivity adjustment coefficient.

3. The adjustable resource flexibility grading assessment method considering duration as described in claim 1, characterized in that, The rule base for classifying adjustable resources based on flexibility thresholds is used to divide adjustable resources into hierarchical regions, obtaining preliminary classification labels for the adjustable resources. Then, confidence checks are performed on these preliminary classification labels to obtain optimized classification labels for the adjustable resources, including: By performing time-series feature matching between the historical response records of adjustable resources and the current operating data, the continuous response duration characteristics of adjustable resources can be obtained. Based on the flexibility threshold partitioning rule base, the continuous response duration feature is hierarchically mapped to obtain the preliminary hierarchical labels of adjustable resources. Extract the resource identifiers from the preliminary classification labels, and perform correlation verification between the resource identifiers and the historical classification records of adjustable resources to obtain the label conflict records of the preliminary classification labels. Perform a hierarchical recalibration operation on the conflicting resources in the tag conflict record to obtain the recalibrated hierarchical tags of the adjustable resources; The recalibrated grading labels are compared with the multidimensional quantitative features of adjustable resources to obtain optimized grading labels for adjustable resources.

4. The adjustable resource flexibility grading assessment method considering duration according to claim 1, characterized in that, The step of generating a scheduling strategy for adjustable resources based on optimized hierarchical labels includes: Based on optimized hierarchical labels, the real-time operation data of adjustable resources is quantified and sorted to obtain the scheduling priority of adjustable resources. Based on scheduling priority, conflict detection is performed on the scheduling period of adjustable resources to obtain the pre-scheduling sequence of adjustable resources. Extract resource identifiers and hierarchical labels from the pre-scheduled sequence, and perform scheduling policy matching on the resource identifiers and hierarchical labels to obtain the initial scheduling policy for the adjustable resources. The consistency of the initial scheduling strategy and the joint response probability distribution characteristics is verified to obtain the verification characterization of the adjustable resources. Based on the verification representation, the initial scheduling policy is modified to obtain the scheduling policy for adjustable resources.

5. A tiered assessment system for adjustable resource flexibility considering duration, characterized in that, include: The multi-mode coupling feature analysis module is used to acquire adjustable resources under multiple modes of integrated power generation, grid-load-storage, and green electricity direct connection within the regional power grid, and to perform multi-mode coupling analysis on the adjustable resources to obtain the joint response probability distribution characteristics of the adjustable resources. This includes: integrating data from the integrated power generation, grid-load-storage, and green electricity direct connection modes within the regional power grid into adjustable resources; performing time-series alignment on the adjustable resources to obtain standardized data; mining dependencies on the standardized data to obtain a multi-dimensional coupling state tensor of the adjustable resources; identifying correlations in each group of the processed standardized data, comprehensively analyzing the interaction and synergistic change patterns between various data points, extracting the mutual influence relationships and linkage states between different resources, and orderly combining and arranging these relationships and linkage states according to a multi-dimensional structure to construct a multi-dimensional coupling state tensor of the adjustable resources that can fully reflect the resource coupling state; performing non-parametric estimation on the adjustable resources based on the multi-dimensional coupling state tensor to obtain the dynamic response statistical characteristics of the adjustable resources; and quantitatively fitting the dynamic response statistical characteristics to obtain the joint response probability distribution characteristics of the adjustable resources. The comprehensive analysis module is used to comprehensively analyze the resource adjustment capability of adjustable resources based on the joint response probability distribution characteristics, obtaining multi-dimensional quantitative characteristics of the adjustable resources. This includes: segmenting the joint response probability distribution characteristics using a variable-length sliding window to obtain local probability segments of the adjustable resources; encoding the local probability segments to obtain the temporal statistical feature spectrum of the adjustable resources; reconstructing the extreme value form of the adjustable indicators based on the temporal statistical feature spectrum to obtain the boundary description of the adjustable resources, including maximum adjustable capacity, longest sustained response time, average ramp rate, and response delay variance; performing vector orthogonal decomposition on the boundary description to obtain independent feature components of the adjustable resources; and spatial mapping of the independent feature components to obtain the multi-dimensional quantitative characteristics of the adjustable resources. The spatial adaptation mapping module is used to construct a planning constraint space for adjustable resources based on power grid safety regulations and multi-mode physical topology constraints, and to map multi-dimensional quantitative features to the planning constraint space to obtain a planning adaptability evaluation representation of adjustable resources. This includes: discretizing and encoding the multi-mode physical topology constraints of adjustable resources to obtain a topological adjacency representation of the adjustable resources; performing semantic vectorization mapping on the power grid safety regulations of the adjustable resources based on the topological adjacency representation to obtain a rule constraint representation of the adjustable resources; performing multi-dimensional tensor dimensionality increase on the rule constraint hyperplane group to obtain a structured weighted basis for the adjustable resources; injecting topological structure into the structured weighted basis to obtain a planning constraint space for the adjustable resources; and performing adaptability analysis on the multi-dimensional quantitative features based on the planning constraint space to obtain a planning adaptability evaluation representation of the adjustable resources. The rule construction module is used to analyze the extreme value distribution patterns of adjustable resources based on the planning adaptability assessment representation, obtain the boundary evolution characteristics of adjustable resources, and construct a rule base for classifying the flexibility threshold of adjustable resources based on the boundary evolution characteristics. This includes: extracting the time-series extreme value trajectory from the planning adaptability assessment representation to obtain extreme state sequence clusters of adjustable resources; and reconstructing the boundary of the multidimensional feature space of adjustable resources based on the extreme state sequence clusters to obtain the boundary evolution characteristics of adjustable resources. Specifically, based on all extreme value point data in the extreme state sequence clusters, the feasible domain of the resources is redefined within the pre-constructed multidimensional feature space, and the extreme value points in each dimension are used as boundary control points. This is achieved through spatial fitting... The formula connects discrete extreme points into a smooth and continuous closed boundary. Simultaneously, it records the positional shifts and morphological changes of the boundary under different operating conditions and time periods, comprehensively presenting the dynamic evolution process of the resource boundary. This forms a boundary evolution feature of adjustable resources that can accurately describe the dynamic changes in the resource's extreme state. Adaptive discretization segmentation of the boundary evolution feature yields effective thresholds for adjustable resources. This process includes tensor decomposition of the boundary evolution feature to obtain local mutation indices for adjustable resources. Specifically, multi-dimensional and multi-level tensor decomposition of the overall boundary evolution feature decomposes the high-dimensional overall boundary feature into multiple low-dimensional... The system identifies the most drastic numerical changes in each local feature component, calculates the magnitude and rate of change at that location, and visually reflects the degree of feature mutation in the local region. This results in a local mutation index for adjustable resources that accurately reflects the intensity of local changes at the boundary. Based on this index, the manifold surface of the boundary evolution features is topologically quantized to obtain a scalar field of complexity for the adjustable resources. Specifically, based on the obtained local mutation index, the manifold surface corresponding to the boundary evolution features is quantized point-by-point according to spatial location. The local mutation values, topological connections, and direction of change at each spatial location are uniformly converted into scalar values. Finally, all scalar values ​​are quantized according to spatial location. The regular arrangement forms a continuous field structure, fully presenting the complex changes and distribution of the boundary as a whole, resulting in a complexity scalar field of adjustable resources that can comprehensively characterize the complexity of boundary changes. Abrupt change point detection filtering is applied to the complexity scalar field to obtain the critical jump vector of the adjustable resources. Specifically, all scalar data within the complexity scalar field are scanned and detected point by point to identify key points where significant numerical jumps occur. Through filtering, interference points and outliers caused by data fluctuations are eliminated, retaining effective points that truly reflect the changes in the critical state of the boundary. These effective points are combined according to their corresponding spatial dimensions and temporal order into directional feature vectors, resulting in the critical jump vector of the adjustable resources that identifies the location and intensity of the critical jump at the boundary.The effective threshold for adjustable resources is obtained by clustering and convergence of critical transition vectors. Specifically, feature similarity is compared among all critical transition vectors, and vectors with similar values ​​and consistent trends are grouped into the same cluster. For each cluster, center calculation and iterative convergence are performed to ensure that each cluster converges stably to a unique center value, representing the standard threshold for the corresponding critical state. This yields a stable, reliable, and practically instructive effective threshold for adjustable resources. The effective threshold is then mapped to a predefined rule space to obtain a rule base for classifying the flexibility threshold of adjustable resources. The hierarchical verification and optimization module is used to divide the adjustable resources into hierarchical regions based on the flexibility threshold rule base, obtain the preliminary hierarchical labels of the adjustable resources, and perform confidence verification on the preliminary hierarchical labels to obtain the optimized hierarchical labels of the adjustable resources. The strategy generation module is used to generate scheduling strategies for adjustable resources based on optimization level labels.

Images