Microgrid cost dredging mechanism design and evaluation method, system, equipment and medium

Abstract

The invention relates to the technical field of power market and public utility management, in particular to a micro-grid cost dredging mechanism design and evaluation method, system, equipment and medium, which comprises the following steps of: assembling a cost conduction rule based on a cost collection unit delimiting structure and a cost conduction diagram; loading life cycle parameters to carry out cash flow prediction under multiple scenes, and generating a multi-scene economy evaluation set; and based on the multi-scene economic evaluation set, performing parameter disturbance, sensitivity and elasticity analysis, identifying key risk factors and generating allocation suggestions, further performing write-back and versioning management on conduction rule parameters, and generating a cost grooming mechanism after iterative correction. The method has the beneficial effect of effectively supporting scientific decision and efficient operation of multiple types of microgrids in a complex market environment.

Description

Technical Field

[0001] This invention relates to the field of electricity market and utility management technology, and in particular to a method, system, equipment and medium for designing and evaluating a microgrid cost mitigation mechanism. Background Technology

[0002] In the field of electricity market and utility management, the current microgrid cost mitigation mechanism mainly revolves around operation scheduling and transaction optimization. It relies on the existing metering and settlement system to collect energy consumption data and combine policy electricity prices and market transaction results to allocate costs and calculate revenues.

[0003] However, existing technologies still have the following problems in practical use: First, market transaction interfaces, external pricing information, and energy consumption metering records belong to different systems, lacking a unified time anchor registration and session number binding mechanism. This leads to difficulties in data alignment across billing periods and a lack of source signature verification, making it impossible to form a complete and credible chain of evidence. Second, relying on manual configuration of the correspondence between assets, users, and contracts is prone to ambiguity, and there is no clear resolution mechanism for conflicts between allocation rules. Third, the results of ex-post multi-scenario economic assessments and allocation factor disturbances lack consistent reference; sensitivity and elasticity analysis cannot effectively identify key risk factors; and rule parameter write-back and version archiving are independent of each other. This not only fails to achieve a closed-loop connection between assessment, write-back, and archiving but also affects the transparency and traceability of cost allocation management. Summary of the Invention

[0004] To solve the above-mentioned technical problems, the present invention provides the following technical solution: In a first aspect, the present invention provides a design and evaluation method for a microgrid cost mitigation mechanism, including unifying data scope and dividing cost unit boundaries based on the functional definition information and entity relationship mapping information of the microgrid, and generating a structured cost collection unit demarcation structure. Based on market transaction data, external pricing information, and energy metering and settlement data, traceable data collection and verification are carried out, and a cost transmission diagram reflecting the flow path and factors of costs between units is constructed according to the cost collection unit boundary structure. Based on the cost collection unit delimitation structure and the cost transmission diagram, cost transmission rules are assembled, and life cycle parameters are loaded to predict cash flow under multiple scenarios, generating a multi-scenario economic evaluation set. Based on the multi-scenario economic assessment set, parameter perturbation, sensitivity and elasticity analysis are performed to identify key risk factors and generate allocation suggestions. Then, the transmission rule parameters are written back and versioned to generate an iteratively corrected cost diversion mechanism.

[0005] As a preferred embodiment of the microgrid cost mitigation mechanism design and evaluation method described in this invention, the step of conducting traceable data collection and verification includes: Read-only economic probes are deployed at the metering, settlement, and market trading ends to collect key data fields; Bind a unique session number to all collected data entries and attach a source signature and time anchor; The collected data is packaged into a traceable cost-benefit observation stream.

[0006] As a preferred embodiment of the microgrid cost transmission mechanism design and evaluation method described in this invention, the cost transmission diagram is constructed by, including, Using the unit identifiers in the cost aggregation unit delimitation structure as nodes and the cost flow relationships in the cost-benefit observation flow as edges, node-edge mapping is performed. By linking the loss allocation factor, capacity allocation factor, and time period allocation factor to the corresponding paths, a cost transmission graph containing a set of nodes, a set of edges, and a factor-path association table is generated.

[0007] As a preferred embodiment of the microgrid cost transmission mechanism design and evaluation method described in this invention, the step of assembling the cost transmission rules includes: Extract the loss allocation factor, capacity allocation factor, and time period allocation factor that are bound to the functional path from the cost transmission diagram. The factors are merged and associated according to nodes and directions to generate assembly rule fragments; When rule conflicts exist within the same time segment, they are pruned according to preset priorities, retaining the higher-priority rules to form a list of propagation rules.

[0008] As a preferred embodiment of the microgrid cost mitigation mechanism design and evaluation method described in this invention, the step of performing parameter perturbation, sensitivity, and resilience analysis includes: Based on the aforementioned multi-scenario economic assessment set, a set of disturbance parameters for electricity price, allocation method and path structure is generated; The perturbation parameters are projected onto the allocation factor level to form an input slice for sensitivity analysis; By calculating the differences and change points of economic indicators before and after the disturbance, the sensitivity and elasticity of single parameters and joint parameters are analyzed.

[0009] As a preferred embodiment of the microgrid cost mitigation mechanism design and evaluation method described in this invention, the generation of the iteratively corrected cost mitigation mechanism includes: The allocation recommendations generated based on key risk factors will be written back to the corresponding parameters in the list of transmission rules in the form of a new version; After writing back the rules, a consistency check and version archiving are performed to form a set of executable cost mitigation solutions that include version reference relationships and switching conditions.

[0010] As a preferred embodiment of the microgrid cost mitigation mechanism design and evaluation method described in this invention, the external pricing information refers to read-only mirror data formed after the structured rate scheme, subsidy items and tiered pricing boundaries from outside the microgrid are version-fixed.

[0011] Secondly, the present invention provides a microgrid cost mitigation mechanism design and evaluation system, comprising: a first generation module, used to perform data standardization and cost unit boundary division based on the functional definition information and entity relationship mapping information of the microgrid, and generate a structured cost collection unit delimitation structure; The module is used to collect and verify traceable data based on market transaction data, external pricing information and energy metering and settlement data, and to construct a cost transmission diagram that reflects the flow path and factors of costs between units according to the cost collection unit boundary structure. The prediction module is used to assemble cost transmission rules based on the cost collection unit boundary structure and the cost transmission diagram, and load life cycle parameters to predict cash flow under multiple scenarios, generating a multi-scenario economic evaluation set. The second generation module is used to perform parameter perturbation, sensitivity and elasticity analysis based on the multi-scenario economic evaluation set, identify key risk factors and generate allocation suggestions, and then write back and version control the transmission rule parameters to generate an iteratively corrected cost diversion mechanism.

[0012] Thirdly, the present invention provides a computer device including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the method described above.

[0013] Fourthly, the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method described above.

[0014] Compared with existing technologies, the beneficial effects of this invention are as follows: In the data acquisition and alignment stage, unified session number identification and precise time anchor registration are adopted to ensure the full traceability and temporal consistency of multi-source heterogeneous data across billing periods; in the rule assembly stage, dynamic filtering based on functional positioning and intelligent pruning of conflict priorities are used to achieve structured organization of transmission rules within time segments and stable connection with economic evaluation; in the evaluation iteration stage, key risk factors are identified based on parameter disturbance and sensitivity analysis, driving closed-loop write-back of rule parameters and version management, forming an auditable and evolvable long-term mechanism. These three aspects are interconnected, connecting the entire lifecycle of "acquisition-assembly-evaluation-optimization", significantly improving the transparency, adaptability and continuous optimization capabilities of microgrid cost allocation. Attached Figure Description

[0015] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 A flowchart illustrating the design and evaluation methods for microgrid cost mitigation mechanisms. Detailed Implementation

[0017] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.

[0018] Example 1, referring to Figure 1 This is the first embodiment of the present invention, which provides a method for designing and evaluating a microgrid cost mitigation mechanism, including: S100: Based on the functional definition information and entity relationship mapping information of the microgrid, data caliber is unified and cost unit boundaries are divided to generate a structured cost collection unit delimitation structure. S200: Based on market transaction data, external pricing information and energy metering and settlement data, traceable data collection and verification are carried out, and a cost transmission diagram reflecting the flow path and factors of costs between units is constructed according to the cost collection unit boundary structure. S300: Based on the cost collection unit delimitation structure and the cost transmission diagram, the cost transmission rules are assembled, and life cycle parameters are loaded to predict cash flow under multiple scenarios, generating a multi-scenario economic evaluation set. S400: Based on the multi-scenario economic assessment set, perform parameter perturbation, sensitivity and elasticity analysis, identify key risk factors and generate allocation suggestions, and then write back and version control the transmission rule parameters to generate an iteratively corrected cost diversion mechanism.

[0019] It should be noted that, in response to the problems of fragmented multi-source data, rigid rule assembly, and disconnected evaluation iteration in existing technologies, this method, through steps S100-S400, first unifies the definitions of multi-source data such as functional positioning and asset-user-contract mapping, clarifies the boundaries of cost collection units, and eliminates mapping ambiguity from the source; then, it constructs a traceable observation link that runs through the entire process of collection-verification-mapping, forming a cost transmission diagram that clearly reflects the cost flow path and allocation factors between units; furthermore, it realizes the dynamic assembly of transmission rules and conflict priority trimming, supporting accurate economic assessment under multiple scenarios; finally, it identifies key risk factors through sensitivity and elasticity analysis, generates parameterized allocation suggestions, and writes them back to the rule system in a closed loop, forming a versioned archive management system.

[0020] Example 2, refer to Figure 1 As an embodiment of the present invention, based on the above embodiment, a method for designing and evaluating a microgrid cost mitigation mechanism is provided.

[0021] In this embodiment of the application, step S100, based on the functional definition information and entity relationship mapping information of the microgrid, unifies data scope and divides cost unit boundaries to generate a structured cost aggregation unit demarcation structure, including the following steps A1-A3: A1: Obtain the functional positioning dictionary and the asset-user-contract three-layer mapping table. Use the field name alignment component to perform mapping consistency verification and identifier binding on the functional positioning dictionary and the three-layer mapping table, and generate the corresponding functional positioning dictionary image and three-layer mapping table image.

[0022] Understandably, the functional positioning dictionary is a standardized template used to uniformly classify and describe various operational scenarios of microgrids (such as supply guarantee type and collaborative type). Its entries record key attributes such as hierarchy, applicable boundaries, and billing-side concerns. The asset-user-contract three-layer mapping table describes the static relationship between physical equipment (assets), electricity users (users), and commercial terms (contracts). Specifically, the asset-user-contract three-layer mapping table uses asset identifier, user identifier, and contract number as the primary key group, and records the scope, unit, time range, and failure conditions with fixed field names.

[0023] It should be noted that the input for this step is the established functional positioning dictionary and the asset-user-contract three-layer mapping table. During implementation, the functional positioning dictionary and the asset-user-contract three-layer mapping table are first loaded into the same data access channel. A field name alignment component is used to perform a consistency comparison of the spelling of entry names, hierarchical identifiers, and primary key groups. A pre-set terminology dictionary is used for alias merging and null value identification. Conflicting entries are registered with source pointers and marked with retention strategies. Subsequently, a mapping consistency check is triggered within the same channel. The mapping base, overlapping intervals, and gap intervals for each asset identifier on the user identifier and contract number are checked. Overlapping entries across lease terms and billing cycles are checked in the boundary conditions, forming a consistency check record. Further, after the consistency check is completed, identifier binding processing is performed. Specifically, functional scenario entries in the functional positioning dictionary are bound to asset identifiers, user identifiers, and contract numbers in the three-layer mapping table in a one-to-many or many-to-one manner. During binding, the binding source, binding priority, and unbinding conditions are written. Conflicting bindings caused by multiple constraints are pruned and registered, forming a binding snapshot. For entries that cannot be verified, anomaly registration is implemented. No valuation substitution is performed; only the missing node, the reason for the missing information, and the suggested filling boundary are recorded to maintain caliber stability. After the above processing is completed, a mirror generation operation is performed within the same session. The functional positioning dictionary and the three-layer mapping table, after consistency verification and identifier binding, are written to the read-only mirror channel, generating a functional positioning dictionary image and a three-layer mapping table image. These images contain a version number, session number, and source signature, and the field names and order are fixed to prevent field drift in subsequent steps. Finally, the output of this step is a functional positioning dictionary image and a three-layer mapping table image, which are stored in pairs using session number and version number, and used as direct input for the next step. The functional positioning dictionary image and the three-layer mapping table image are passed to the input position of A2 for its use. Simultaneously, in the cross-main step reference relationship, these images have bypass read permissions for the subsequent S200 economic probe configuration and a read-only reference relationship for the S300 functional positioning assembly.

[0024] A2: Based on the aforementioned image, perform unified processing of measurement and settlement standards and unit boundary verification to generate a multi-level cost unit list.

[0025] It should be noted that the input for this step is the aforementioned functional positioning dictionary mirror and three-layer mapping table mirror. During implementation, the functional positioning dictionary mirror is first invoked within the same session to read the functional scenario item sets corresponding to supply guarantee and collaborative types. Then, the asset identifier and user identifier matching the item sets are searched in the three-layer mapping table mirror to generate an asset list and a user group set. The asset list describes the aggregation of equipment-level objects, including fixed field names such as equipment category, installation loop, geographical location information, and maintenance affiliation. The user group set describes the grouping of user-level objects, including fixed field names such as group boundaries, billing relationships, and historical change markers. Subsequently, a unified metering-settlement caliber processing is performed. Specifically, the caliber, dimension, and time range fields in the three-layer mapping table mirror are read. Dimension conversion and sampling time benchmark alignment are performed on metering items from different data sources. The rate version, discount strategy, and tiered boundary of the settlement items are registered on the same timeline. A mapping relationship between metering fields and settlement fields is established, and the mapping relationship is created as a queryable mapping snapshot. After standardizing the definitions, the execution unit boundary verification is performed. Using the boundary descriptions for supply-supporting and collaborative types in the functional positioning dictionary mirror, a hierarchical traversal is conducted at the equipment, loop, and user group levels. Conflicts are identified for overlapping boundaries across levels, gaps are registered for uncovered areas, and duplicate inclusions are pruned. The pruning record includes the pruning reason, pruning time, and pruning strategy. Further, the asset list and user group set, after standardization and boundary verification, are merged and summarized to form a multi-level cost unit candidate set. Hierarchical identifiers, attribution paths, and metering-settlement mapping pointers are added to the candidate set. A multi-level cost unit list is generated using stable field order, and the correspondence between the list entries and the functional positioning dictionary mirror is recorded. During the above processing, abnormal entries are only marked and their source recorded, without numerical substitution. For list entries that have undergone cross-period changes, the mapping pointers and effective boundaries of the two versions before and after the change are recorded to maintain consistency in subsequent steps. Finally, the output of this step is a multi-level cost unit list, which contains entries and boundary descriptions at three levels: device level, loop level, and user group level, and is archived with the same session number and version number. This output is passed to the input position of A3 for it to call. At the same time, the list serves as a node construction reference for S200 and a bypass read object for the path filtering pre-set of S300 in the cross-main step.

[0026] A3: Perform hierarchical labeling and cross-unit association registration on the multi-level cost unit list to generate the cost collection unit boundary structure.

[0027] It should be noted that the input for this step is the aforementioned multi-level cost unit list. During implementation, firstly, within the same session, equipment-level, loop-level, and user group-level entries in the multi-level cost unit list are read. Based on the fixed hierarchical identifiers and attribution paths in the list, each level of entry is labeled with a hierarchical tag. The labeling includes hierarchical labels, upstream paths, downstream paths, and metering-settlement mapping pointers. Historically changed entries are also appended with change segment identifiers and effective boundaries to ensure the stability of hierarchical labeling within the session. Subsequently, cross-unit association registration is performed, establishing associations between entries at different levels but with business connections. Association relationships include three fixed field names: association direction, association strength, and association system source. During association registration, if a one-to-many or many-to-one cross-mapping occurs between the equipment level and the loop level, a pruning strategy and a priority reading flag are written into the association record, and a separate pruning record is created for conflicting associations. Furthermore, the list items that have completed hierarchical labeling and cross-unit association registration are delimited and aggregated. Based on the boundary descriptions of supply-guarantee and collaborative types in the functional positioning dictionary mirror, multi-level items are merged in the order of "equipment—loop—user group," ensuring that each cost collection unit has a single upstream path and enumerable downstream paths. The corresponding metering-settlement mapping pointer and association registration pointer are written under the unit item. For units spanning multiple periods and boundaries, parallel delimitation fragments are generated, recording the effective boundary, association direction, and attribution strategy of each fragment. After delimitation and aggregation, a cost collection unit delimitation structure is generated. This structure records the unit identifier, hierarchical label, upstream path, downstream path, metering-settlement mapping pointer, association registration pointer, and version number using unified field names. A session number and source signature are written in the structure header for easy read-only referencing in subsequent steps. Finally, the output of this step is the cost aggregation unit delimitation structure. The cost aggregation unit delimitation structure serves as one of the inputs to the subsequent S200, which is used to provide node boundaries and belonging paths. It also serves as one of the inputs to S300, which is used for unit boundary references before functional positioning assembly and path filtering.

[0028] Preferably, this step forms a stable cost aggregation unit boundary structure through hierarchical labeling, cross-unit association, and boundary aggregation, providing directly referable boundary and path information for subsequent construction of cost transmission relationships and assembly of differentiated transmission rules.

[0029] In an optional implementation, the generation of the structured cost aggregation unit delimitation structure in step S100 can also be achieved through an adaptive unit delimitation method based on cluster analysis. That is, by using historical energy consumption metering data and equipment operating parameters, a density clustering algorithm is used to perform similarity analysis on the load characteristics and spatiotemporal coupling degree of each physical node and user in the microgrid. Objects with similar load curves and close electrical distances are automatically and dynamically aggregated into preliminary cost units. Then, the supply guarantee / coordination boundary constraints and asset-user-contract mapping relationship in the functional positioning dictionary are combined for correction and trimming, and finally a hierarchical cost aggregation unit delimitation structure with dynamic weight labels and flexible boundary descriptions is generated.

[0030] In another optional implementation, the generation of the structured cost collection unit delimitation structure in step S100 can also be achieved through a graph-based topology-driven delimitation method. This involves abstracting the microgrid physical topology and business relationships into an attribute graph model, using devices, users, and contracts as vertices, and power flow direction, settlement relationship, and attribution path as weighted edges. A graph community discovery algorithm is used to automatically identify strongly correlated node clusters as candidate sets of cost collection units. Then, based on the priority of functional positioning paths and the betweenness centrality between nodes, unit boundaries are divided and cross-unit associations are labeled, outputting a structured cost unit delimitation structure that includes topological order relationships, path dependency strength, and node importance scores.

[0031] In this embodiment of the application, step S200 involves traceable data collection and verification based on market transaction data, external pricing information, and energy consumption metering and settlement data. Based on the cost aggregation unit demarcation structure, a cost transmission diagram reflecting the cost flow path and factors between units is constructed, including the following steps B1-B3: B1: Acquire market transaction data, external pricing information, and energy metering and settlement data. Collect data through read-only economic probes deployed at the metering, settlement, and market transaction ends, and bind all collected data to a unified session number, time anchor, and source signature to generate a traceable cost-benefit observation stream.

[0032] Understandably, an economic probe is a technical component used for data collection. Its architecture is based on a non-intrusive design and it primarily collects data from the metering, settlement, and transaction ends.

[0033] It should be noted that the input for this step is a mirror image of the cost collection unit's boundary structure and functional positioning dictionary. Within the same session, it calls the market transaction interface and external pricing information, while simultaneously accessing energy consumption metering and settlement records. Specifically, the market transaction interface refers to the data interaction port open to the microgrid trading platform, containing fixed field names such as matching results, order records, transaction price, transaction volume, quotation period, and transaction channel identifier. The set of fields exposed by the interface and access permissions are controlled by the gateway on the trading platform side. External pricing information refers to a read-only mirror image formed after structuring and versioning external electricity price text, rate schemes, subsidy items, peak-valley boundaries, and tiered segments. The mirror image records fields such as version number, source pointer, release time, applicable area, applicable object, and failure conditions. Energy consumption metering and settlement records refer to a structured data set from the metering acquisition system and the settlement system. The metering part includes fields such as time series sample values, metering point number, sampling period, and quality markers, while the settlement part includes fields such as payment period, rate version, discount strategy, pre-verification result, actual settlement amount, and payment record. During implementation, the system first completes the access and session initialization processing for the three types of inputs on the same acquisition channel, creating a new session number and registering the creation time, caller identifier, source signature, and permission scope. Then, the economic probe deployment process is initiated. The economic probe is defined as a non-intrusive acquisition strategy targeting the metering, settlement, and trading ends. The probe only reads necessary fields and writes the session number, without writing back to the source system or triggering any state changes in the source system. Probe deployment follows a three-layer strategy: the first layer targets the metering acquisition port, binding the metering point number, sampling period, and quality marker; the second layer targets the settlement port, binding the payment period, rate version, and discount strategy; and the third layer targets the market trading port, binding the transaction price, transaction volume, time period identifier, and channel identifier. For each probe trigger action, the system records the probe trigger time, trigger source, number of field readings, exception code, and number of retries, forming a probe trigger log (it is important to emphasize that this log-based acquisition reduces the load and risk on the source system). Furthermore, session number binding is implemented throughout the entire data collection process for all three types of inputs. All collected field values ​​are written to the same session number (ideally, this isolation mechanism limits data processing to the scope of the current session, decoupling it from the source system and avoiding cross-system dependencies or conflicts). If cross-period data is encountered, a sub-session number is created under the same main session number, recording the inheritance relationship between the sub-session and the main session. Regarding exception handling, if a field is missing, timestamp is duplicated, permission is denied, or network timeout occurs in a certain collection window, the system marks the entry as missing, does not perform a replacement, and only records the reason for the missing entry, the missing period, and the corresponding source port (ideally, this ensures that the data collection process does not actively modify or infer data due to data problems, maintaining the originality of the source data).After completing the above actions, the system enters the observation construction phase. Fields from measurement, settlement, and transactions are concatenated according to session number and time sequence to form a time-series ordered observation entry stream. A source signature and session number are appended to the header of each entry. The entry body contains three sets of fields: measurement fields, rate fields, and transaction fields. The tail of each entry records the reading strategy and quality markers. Finally, the output of this step is a traceable cost-benefit observation stream, with the output field named "Traceable Cost-Benefit Observation Stream." This observation stream enters the input position of subsequent step B2 for field extraction, time anchor registration, and source signature verification. Simultaneously, at the cross-main step level, it provides a bypass read-only reference for subsequent differential transmission rule assembly and multi-scenario economic assessment.

[0034] Specifically, to further quantify the data acquisition efficiency triggered by the probe, a cross-correlation function is introduced to evaluate the alignment of different source data over time. Formula ① is defined as follows: In the formula: C 1 (τ) represents the cross-correlation value at time shift τ, used to measure the alignment consistency between the measurement field and the settlement field on the time axis; x 1 (t) represents the time-series sampled values ​​extracted from energy consumption metering records, sourced from sampled values ​​bound to metering point numbers; y 1 (t) represents the actual settlement amount sequence extracted from the settlement record, which is the actual settlement amount bound to the payment period; t represents the sampling time index, which takes the value of an integer; τ represents the time shift, which is preset at the system level according to the business time characteristics of the data source, and takes the value of an integer; T represents the length of the time series.

[0035] It is understandable that time-series sampled values ​​are extracted from energy consumption metering records and denoted as x. 1 (t), the actual settlement amount sequence is extracted from the settlement record and denoted as y. 1 (t), together they form the cross-correlation sequence C in formula ①. 1 (τ).

[0036] Furthermore, C in formula ① 1 (τ) serves as the input for subsequent time anchor alignment, used to identify the optimal time shift. Subsequently, to optimize the data integrity of session number binding, a resource conflict constraint model based on linear programming is introduced. Formula ② is defined as follows: In the formula: z 1 (i a ) shows the i-th a The actual field values ​​of each data collection point; Indicates the expected field value; A represents the constraint matrix; b represents the constraint vector; N represents the number of data collection points; i a : indicates the i-th a Index of each data collection point; z 1 : indicates that it is composed of all z 1 (i a A vector composed of ).

[0037] It is understandable that the transaction price sequence is extracted from the market trading interface and denoted as z. 1 (i a The rate version is extracted from external pricing information and recorded as... In Formula ②, the two factors are minimized through linear programming to ensure that data acquisition satisfies resource constraints under session number binding. The cross-correlation sequence C in Formula ①... 1 (τ) is used to construct the constraints in Formula ②, aligning the time axis by identifying the optimal τ. The output of this section is the initially aligned acquired data field, which is then used as input for the field extraction operation in B2.

[0038] Furthermore, following the aforementioned cross-correlation sequence C 1 (τ) and the linear programming result z 1 (i a The system then proceeds to time anchor registration and source signature verification. A time anchor is defined as a set of reference points used to unify time coordinates from different sources, encompassing three levels: acquisition time, payment period time, and market matching time. The registration process reads the sampling time and transaction period of the observation entries and unifies them to the session timeline. To accurately align multi-source time series, a Short-Time Fourier Transform (STFT) is introduced to extract local time frequency features. Formula ③ is defined as follows: In the formula: X 2 (t b f) represents time point t b STFT coefficients at frequency f; x 2 (n) represents the input time series; w(n−t) b ) represents the window function; t b The time anchor index is represented by f; the frequency is represented by n; the time index is represented by t. b : Represents the time anchor index; n: Represents the time index; j: Represents the imaginary unit; π: Pi; exp: Natural exponential function.

[0039] It is understandable that the sampled value is extracted from the energy consumption metering record and recorded as x. 2 (n), from formula ① C 1 (τ) Determine the resampling parameters, X in formula ③ 2 (t bf) is used to identify frequency consistency within a time segment. The specific type and length of the window function are preset algorithm parameters based on the typical characteristics of the time series to be aligned. Selection logic: STFT is used for time alignment, and the selection of the window function needs to strike a balance between time resolution and frequency resolution. Specific selection: Window type: Commonly used window functions include Hanning or Hamming windows to reduce spectral leakage. The specific type is preset by the system; Window length: The window length (i.e., the range of n in the formula) should match the length of key business cycles in the sequence (such as electricity price peak / valley cycles, settlement cycles) to ensure that business-meaning frequency components are captured. This is a system parameter preset based on business knowledge.

[0040] Furthermore, to verify the integrity of the source signature, an attention-weighted scoring mechanism is introduced. Formula ④ is defined as follows: In the formula: α 2 (k c ) represents the k-th c Attention weights for each source signature; s 2 (k c ) represents the signature similarity score; β represents the temperature parameter; M represents the number of signatures; k c : indicates the k-th c An index of source signatures; j c : Indicates the summation index.

[0041] It is understandable that the source pointer is extracted from external pricing information and denoted as s. 2 (k c ), α in formula ④ 2 (k c The X in formula ③ is used for weighted verification results. 2 (t b f) serves as the time alignment reference input formula ④, which aligns the signature timestamp through a window function. The output of this section consists of the observation entries after time anchor registration, which are inputs for time anchor registration and source signature verification in B2.

[0042] Furthermore, following the aforementioned STFT coefficient X 2 (t b f) and attention weight α 2 (k c The system further executes the observation construction phase, concatenating fields from metering, settlement, and transaction based on session number and time sequence to form a stream of observation entries ordered by time series. To ensure the quality of the stream's structured encapsulation, time-frequency feature fusion based on wavelet packet energy is introduced. Formula ⑤ is defined as follows: In the formula: E 3 (m d ) represents the m-th d Wavelet packet energy of a time segment; W 3 (m d ,f d ) represents wavelet packet coefficients; m d Indicates the time segment index; f d Indicates a frequency sub-band; F represents the frequency set; Modular square function; : Belongs to operators.

[0043] It is understandable that W is obtained by concatenating the fields from the energy consumption metering record and the market transaction interface. 3 (m d ,f d E in formula ⑤ 3 (m d This is used to quantify the time-frequency stability of observation entries.

[0044] Finally, to generate a traceable cost-benefit observation stream, a logical consistency scoring function is introduced. Formula ⑥ is defined as follows: In the formula: L 4 Indicates logical consistency score; o 3 (p e ) represents the p-th e The actual value of each observation entry; p e : indicates the p-th e An index of observation entries; δ(⋅) represents the expected value; δ(⋅) represents the indicator function, which is 1 when the parameters are equal and 0 otherwise; P represents the number of entries.

[0045] It is understandable that E from formula ⑤ 3 (m d After normalization, it is used as o 3 (p e The weight of ) and L in formula ⑥ 4 Used to label the quality of the observed flow. E in Equation ⑤ 3 (m d Enter the consistency calculation of formula ⑥.

[0046] B2: Extract fields, register time anchors, and verify source signatures for the cost-benefit observation stream to generate an observation session record set.

[0047] It should be noted that the input for this step is a traceable cost-benefit observation stream, referencing the unit identifier and path boundary recorded in the cost aggregation unit demarcation structure within the same session. During implementation, the traceable cost-benefit observation stream first undergoes field extraction, limited to three sets: measurement fields, rate fields, and transaction fields. Measurement fields include measurement point number, sampling time, sample value, quality marker, and sampling period; rate fields include rate version, time period classification, tiered segment, subsidy item, discount strategy, and tax item; transaction fields include transaction price, transaction volume, matching number, transaction channel identifier, and transaction time period. Field extraction employs a fixed field name mapping rule and a field missing tolerance strategy (ensuring lightweight data collection). If a field is missing, the system writes the missing code and source pointer to the record without interpolation or inference, and without altering the original numerical expression. Subsequently, the time anchor registration process begins. Time anchors are defined as a set of reference points used to unify time coordinates from different sources, encompassing three levels: collection time, payment period, and market matching time. The registration process first reads the sampling time and transaction period of the observation items, unifies them to the session timeline, and then writes the period anchors for items that cross the payment period boundary, recording the start and end of the payment period, the payment period number, and the settlement deadline. If the observation items have out-of-order timestamps or duplicate timestamps, the system marks the item as an out-of-order item while preserving the original order, using an out-of-order identifier field for labeling. After completing the time anchor registration, source signature verification is performed. The source signature refers to the integrity certificate of the source of the observation item. The verification step reads the source signature and source pointer in the item and compares them with the session registration information and the source system signature disclosure record. If the signature does not match or is missing, the system marks the item as a signature exception item, records the exception type and a suggested resampling window, without replacing or correcting the item content. Furthermore, based on the unit identifier and path boundary in the cost collection unit demarcation structure, the system associates the items that have completed time anchor registration and source signature verification with the corresponding units and paths, forming a binding relationship between items and units, and records the binding time, the rule, and the reading priority in the binding relationship. Therefore, the system generates an observation session record set, which is a collection of entries organized by session number. Each entry in the set contains three sets of content: a measurement field, a rate field, and a transaction field, as well as metadata fields related to time anchor, source signature, and unit binding. Finally, the output of this step is the observation session record set, with the output field named "Observation Session Record Set." This output is used in the subsequent input position B3 for node-edge mapping and path factor attachment. Simultaneously, this output provides queryable entry source records for differentiated transmission rule assembly and multi-scenario economic assessment in the main step chain, facilitating rule filtering and scenario loading in subsequent steps.

[0048] B3: Based on the unit identifier and path information in the cost collection unit delimitation structure, the observation session record set is mapped to nodes and edges, and loss sharing factor, capacity sharing factor and time period sharing factor are attached to generate a cost transmission graph containing a set of nodes, a set of edges and a factor-path association table.

[0049] It should be noted that the input for this step is the observation session record set, referencing the upstream and downstream path descriptions in the cost aggregation unit delimitation structure within the same session. During implementation, a node set is first constructed. This set originates from the unit identifier (the unique identifier of each node) in the cost aggregation unit delimitation structure. Nodes are categorized into three types: equipment-level nodes, loop-level nodes, and user group-level nodes. Each type of node contains fixed field names such as a hierarchical label (identifying whether the node belongs to the equipment, loop, or user group level), upstream path (pointing to the upstream source path of the node), downstream path (pointing to the downstream destination path of the node), metering-settlement mapping pointer (pointing to the mapping relationship between metering and settlement data related to the node), and association registration pointer (pointing to the cross-unit association records between the node and other nodes). Following this, a node-edge mapping process is initiated. Entries in the observation session record set are merged into corresponding nodes according to their unit binding relationships. Edges are then created based on upstream and downstream paths. Each edge records fields such as the starting node, ending node, associated path (the cost transmission path to which the edge belongs), session number (the session identifier that created the edge, used for traceability), time period range (the effective start and end times of the edge), and number of entries (the number of observation entries merged onto the edge). Edge creation follows the path order and priority strategy given by the delimitation structure, avoiding crossing undefined path boundaries and preventing the generation of unregistered paths. If an observation entry matches multiple nodes during the binding process, the system makes a decision based on the priority reading flag in the association registration. The decision result is written to the conflict pruning record, retaining the source pointer and time period range of the unused branch. After completing the node-edge mapping, the path factor attachment process begins. Path factors include three main categories: loss sharing factors, capacity sharing factors, and time period sharing factors. During attachment, entries in the functional positioning dictionary mirror are read, and different factors are attached to the corresponding paths according to the definitions of supply guarantee and collaborative types. Loss allocation factors describe the rules for energy flow loss allocation and accounting channels along the path; capacity allocation factors describe the rules for allocating capacity-related costs along the path and their constraint boundaries; and time-based allocation factors describe the rules for cost allocation during different time periods such as peak, flat, and valley periods. The hooking action is executed by a rule assembly engine. The engine reads the factor dictionary, path description, and time anchor information from the observation session record set to generate a factor-path association table (this is the core storage table for path factors, recording the binding relationship between factors and paths). If a path lacks factor configuration, the engine registers the missing status and suggested completion items in the association table, without performing inference or estimation to fill in the gaps. Furthermore, the system marks the effective scope and conflict priority of path factors based on the time period range and number of entries in the edge records, forming a set of effective fragments for path factors. If multiple factors conflict in a certain time fragment, the system performs pruning according to a preset conflict priority strategy. The pruning result is written to the factor pruning record, and the source and version identifiers of the unused factors are retained.After the node-edge mapping and path factor association are completed, the system generates a cost transmission graph. This graph uses a session-oriented structured representation and includes fixed field names such as node set, edge set, factor-path association table, effective fragment set, conflict pruning record, and version information. Finally, the output of this step is the cost transmission graph, with the output field name "Cost Transmission Graph." This output is used in the subsequent C1 input position for function positioning and path filtering. Simultaneously, this output serves as the structured basis for subsequent transmission rule assembly and cash flow forecasting in the main step chain, supporting differentiated transmission rule assembly and lifecycle parameter mirror loading for calculation.

[0050] Preferably, this step involves mapping nodes to edges and attaching path factors to the observation session record set to form a cost transmission graph with a clear structure, traceable source, and well-defined factor effective scope. This provides a stable structure and queryable metadata for subsequent functional positioning path selection and transmission rule assembly.

[0051] In an optional implementation, the cost transmission diagram reflecting the flow path of costs and factors between units in step S200 can also be constructed through a blockchain-based distributed ledger traceability method. This involves storing the matching records of market transaction interfaces, the version snapshot of policy electricity price mirrors, and energy consumption metering settlement data on the blockchain through hashing. Each participating node jointly maintains an immutable distributed ledger, and smart contracts are used to automatically execute the node-edge mapping logic and factor attachment rules between cost collection units. Each cost flow is verified for its source through on-chain timestamps and digital signatures, ultimately generating a cost transmission diagram with multi-party consensus and full auditability.

[0052] In another optional implementation, the cost transmission diagram reflecting the flow path and factors of costs between units in step S200 can also be constructed through dynamic extrapolation modeling based on digital twin simulation. That is, a digital twin of the physical entity and business relationship of the microgrid is constructed, and metering and transaction data are accessed in real time as simulation input. Through virtual operation, the flow trajectory and loss distribution of costs between equipment, circuits and user units under different operating conditions are extrapolated. Key transmission paths are dynamically identified and the weight allocation of time period and capacity factor is adaptively adjusted. After the simulation results are compared and verified with the measured data, a high-fidelity and predictable cost transmission diagram is generated.

[0053] In this embodiment of the application, step S300 involves assembling cost transmission rules based on the cost collection unit delimitation structure and the cost transmission diagram, and loading lifecycle parameters to perform cash flow forecasting under multiple scenarios, generating a multi-scenario economic evaluation set, including the following steps C1-C3: C1: Based on the cost collection unit delimitation structure and the cost transmission diagram, a set of functional positioning paths is generated through functional positioning assembly and path filtering processing.

[0054] It should be noted that the input for this step is the cost aggregation unit boundary structure and the cost transmission diagram, and the set of entries in the functional positioning dictionary mirror is used as the basis for assembly within the same session. Specifically, the aforementioned cost aggregation unit boundary structure serves as the authoritative reference for unit boundaries and upstream and downstream paths, and the aforementioned cost transmission diagram serves as the structured carrier of node-edge relationships and factor effective fragments. At the same time, supply-guarantee and collaborative entries from the functional positioning dictionary mirror are read into the session registration. The entries include functional labels, applicable boundaries, and priority descriptions. At the start of processing, the system establishes a functional assembly context within the assembly channel, reads the hierarchical labels, upstream path, and downstream path fields from the cost aggregation unit boundary structure, and compares them one by one with the node set and edge set in the cost transmission diagram to form a list of assembleable nodes and a list of candidate paths. If a node is missing or a path is not registered during the comparison process, the system only records the missing status and source pointer in the assembly context without making any supplementary values ​​or replacements. This triggers a function positioning and assembly action. The assembly action matches the function tags in the function positioning dictionary mirror with the list of assembleable nodes. The matching rules include three categories of judgments: scenario adaptation, path constraints, and priority pruning. Scenario adaptation determines whether a node belongs to the supply-guarantee or collaborative application boundary. Path constraints determine whether a candidate path is within the allowed path set of the function entry. Priority pruning prunes paths based on the priority relationship given by the function entry when multiple paths exist. For node-path pairs that satisfy scenario adaptation and path constraints, the system generates an assembly record in the assembly context. The record includes the node identifier, path identifier, function tag, and pruned priority marker. For candidates that do not meet the constraints, the system registers the reason for removal and the source pointer in the assembly context, preserving traceability. After assembly is complete, the system enters the path filtering process. Path filtering performs time period and factor weighting judgment on the edge set, reads the factor-path association table and factor effective segment set from the cost transmission diagram, expands each candidate path on the session time axis, and checks whether the loss sharing factor, capacity sharing factor, and time period sharing factor related to the function tag have effective segments, and selects overlapping segments according to the priority markers of the assembly record. If multiple paths in the same time segment meet the assembly conditions, the system selects the first path in descending order of priority and writes the identifier of the discarded path and the discarded time segment into the filtering log; if no path meets the assembly conditions, the system records the time segment as a gap segment and processes it through a bypass strategy in subsequent rule assembly. After filtering is complete, the system merges the effective paths in each time segment, generates a continuous path sequence oriented towards the function tag, and writes switching instructions for path switching across segments so that subsequent rule assembly can identify boundary changes.Finally, the system generates a set of functional location paths within the session. This set is defined as "a set of path sequences organized by functional labels after functional assembly and path filtering." Each entry contains fixed field names such as functional label, node sequence, path sequence, time segment, priority marker, and switching indicator. The output of this step is the set of functional location paths, with the output field name "Functional Location Path Set." This output is naturally input into subsequent step C2 and serves as the direct basis for the assembly of differentiated transmission rules and the source of the path skeleton for multi-scenario economic evaluation in the cross-main step reference relationship. Simultaneously, it maintains a read-only association with the aforementioned cost collection unit boundary structure and cost transmission diagram, allowing subsequent steps to verify and trace it by session number.

[0055] C2: Extract loss sharing factor, capacity sharing factor and time period sharing factor from the functional positioning path set, perform transmission rule assembly and conflict priority pruning, and generate target transmission scheme package.

[0056] It should be noted that the input for this step is a set of functional positioning paths, and within the same session, it reads only the factor-path association table and factor effective segment set in the cost transmission diagram. Specifically, the sequence of entries of the functional positioning path set is loaded in the rule assembly channel, and the expansion result of each path on the time segment is used as an assembly unit. Then, the three types of configurations bound to the path—loss allocation factor, capacity allocation factor, and time period allocation factor—are read from the factor-path association table of the cost transmission diagram. The fixed field names of the three types of configurations record the factor name, applicable node, applicable direction, allocation caliber, accounting channel, and effective scope, respectively. If a factor is missing or the effective scope is blank during the reading process, the system does not make inferences but directly registers the factor missing status and suggested completion items on the assembly unit. After factor extraction is completed, the system performs transmission rule assembly on the assembly unit. Transmission rule assembly is defined as "path-based transmission and allocation description formed under the joint constraints of functional positioning path and three types of allocation factors". The assembly action merges loss allocation factors and capacity allocation factors according to nodes and directions, and segments time-period allocation factors according to time segments and associates them with the aforementioned merging results piece by piece to generate assembly rule fragments. During the merging process, if multiple loss allocation factors appear in the same direction at the same node, the system selects them according to the preset priority labels of the factors and writes the source and version of the covered factors in the assembly log. If there is a boundary inconsistency between capacity allocation factors and time-period allocation factors, the system uses the switching indication in the functional positioning path set as a constraint to split the conflicting fragments into two adjacent fragments and assemble them separately, thereby maintaining the accurate separation of the effective scope without introducing valuation. After the assembly rule fragments are generated, the system enters the conflict priority pruning stage. This stage reads the priority tags and source levels from the assembly rule fragments and prunes multiple sets of rules within the same time segment according to a pre-defined priority relationship table. The set of rules with higher priority and source level is retained, and the pruned rules are written to a pruning record, retaining the source pointer and applicable fragment for subsequent review. The results of assembly and pruning are merged into a transmission rule list within the assembly channel. This list is a set of rules organized by function tags, paths, and time segments within the session. Each entry contains fixed field names such as node pairs, direction markers, loss allocation caliber, capacity allocation caliber, time period allocation configuration, and pruning record pointers. Subsequently, the system generates a target transmission scheme package based on the transmission rule list. This target transmission scheme package is defined as a "path-based transmission scheme that can be directly invoked in subsequent economic evaluations." Its internal structure includes fields such as function tags, path indexes, assembly rule fragment sets, conflict priority pruning records, missing factor registration, and session number. The package header registers the version number and source signature.Ultimately, the output of this step is the target transmission scheme package, with the output field named "target transmission scheme package". This output is naturally input into the "target transmission scheme package" of the subsequent step C3, and becomes the sole source of rules for multi-scenario economic evaluation at the cross-main step level. At the same time, a backtracking pointer is established for the aforementioned functional positioning path set to maintain structural consistency and reading continuity between assembly and evaluation.

[0057] It should be further noted that the priority table is authoritative and pre-defined, its core source being the business rules defined in the functional positioning dictionary. Specifically: 1. Fundamental Source: Functional Positioning Dictionary Priority relationships are primarily based on the inherent priority descriptions of different functional scenarios, such as "supply guarantee" and "collaboration," in the functional positioning dictionary. For example, supply guarantee functions typically have a higher priority than collaborative functions.

[0058] 2. Direct input: Function location path set During rule assembly, the system reads the generated "function location path set". The entries in this set already contain "priority flags" obtained through function location assembly. These flags are the result of directly applying the function location dictionary rules.

[0059] 3. Rule carrier: Transmitting rule fragments The "assembly rule fragment" generated during the assembly process inherits the above priority. It contains two key fields: "priority label" and "source level". Together, they form the direct basis for the priority table to make judgments.

[0060] Furthermore, the core features of the priority table maintenance method are "pre-set" and "mirroring", emphasizing stability and traceability, rather than dynamic modification in the session.

[0061] 1. Preset and Fixed: This table is a set of baseline rules preset during system initialization, and is not dynamically generated or modified during cost allocation calculation. It is fixed in the system as a basic configuration.

[0062] 2. Mirrored reference: When needed, the system reads its read-only mirror (such as the function location dictionary mirror) to ensure that priority rules remain consistent throughout the session and do not drift.

[0063] 3. Version Management: If the priority table needs to be updated due to changes in business rules, a new version will be created. After the new version takes effect, the old version will be retained as an archived version to ensure the repeatability and auditability of historical sessions. Within a single session, the priority table remains unchanged.

[0064] Overall, the priority table is a pre-defined, multi-dimensional decision matrix. It takes the "priority label" (e.g., functional scenario = supply guarantee type) and "source level" of each rule fragment as input, and outputs a comprehensive priority order based on pre-defined rules (e.g., supply guarantee type > collaborative type; national policy > local policy), thereby guiding the system to make deterministic choices when conflicts occur.

[0065] Specifically, to handle dependencies in the path sequence, a topology sorting tool is introduced to ensure the consistency of the path assembly order. Formula ⑦ is defined as follows: In the formula: O 7 G represents the topological sorting result of the path sequence; 7 This represents a path dependency graph; TopoSort represents the topological sorting function, a standard graph algorithm tool.

[0066] It is understandable that the node sequence and path sequence are extracted from the functional location path set and denoted as G. 7 O in formula ⑦ 7 As the basis for the path processing order in the assembly channel, the path dependency graph G 7 The edge weights are determined by the path attributes (such as priority markers) in the functional positioning path set and the factor relationships in the cost transmission diagram. Definition method: Edge weights are used to quantify the dependence strength between paths. Specific content: Weights may be based on: (1) Business priority: The "priority markers" generated in the functional positioning assembly can be used directly or after conversion as weights; (2) Data flow strength: Based on the "number of entries" of the edges in the cost transmission diagram or the scale of associated cash flow; (3) Logical correlation: The strength of the logical relationship between paths due to shared nodes or factors. The specific calculation rules for weights are part of the system's preset mapping strategy.

[0067] Furthermore, to optimize resource allocation during the factor merging process, a linear programming model is introduced to minimize caliber bias. Formula ⑧ is defined as follows: In the formula: f 7 (i a ) represents the i-th a The loss amortization factor value of each node; i a : indicates the i-th a The index of each node; g 7 (i a ) represents the i-th a The capacity allocation factor value of each node; A 7 Represents the constraint matrix; b 7 Represents the constraint vector; x 7 This represents the vector of decision variables; N represents the number of nodes.

[0068] It is understandable that the loss allocation factor is extracted from the cost transmission diagram and denoted as f. 7 (i a The capacity allocation factor is denoted as g. 7 (i a Formula ⑧ minimizes the difference in scope through linear programming, but is constrained by time boundary constraints. Formula ⑦ has O... 7 The node index sorting used in Formula ⑧ ensures consistent processing order. The output of this section is a preliminary assembly rule fragment, containing node pairs, direction markers, and allocation calibers, with the input pruned by the conflict priority of C2.

[0069] Furthermore, following the aforementioned topological sorting result O 7 and linear programming solution x 7 The system then enters the conflict priority pruning phase. Conflict priority pruning reads the priority tags and source levels from the assembly rule fragments and prunes multiple sets of rules within the same time fragment according to a pre-defined priority relationship table. To quantify factor priority, an attention-weighted scoring mechanism is introduced. Formula 9 is defined as follows: In the formula: α 9 (k b ) represents the k-th b Attention weights of each factor; s 9 (k b ) represents the k-th b Priority label scores for each factor; k b : indicates the k-th b Index of factors; β 9 Represents the temperature parameter; M represents the number of factors; j b : Indicates the summation index.

[0070] It is understandable that the priority label is extracted from the assembly rule fragment and denoted as s. 9 (k b ), α in formula ⑨ 9 (k b ) is used for weighting factor priority.

[0071] Furthermore, to handle rule conflicts within a time segment, a pruning strategy under time window constraints is introduced. Formula 10 is defined as follows: In the formula: C 10 c: represents the optimal set of pruning rules; c: represents candidate rules; t represents the set of candidate rules; c This represents the time segment index; T represents the total number of time segments. This indicates the indicator function, when rule c is at time t. c It is 1 when it is effective, otherwise it is 0; α 9 (c) represents the attention weight of rule c.

[0072] It is understandable that α from formula ⑨ 9 (k b As input, formula 10 selects the optimal rule by maximizing the weighted effective time. The α in formula 19... 9 (k b This is directly used for the scoring calculation in Formula 10. The output of this section is a list of transmission rules, including node pairs, direction markers, allocation calibers, and clipping record pointers, which are mirrored and loaded by the life cycle parameters of C3 and the input for cash flow forecast calculation.

[0073] Furthermore, following the aforementioned attention weight α 9 (k b ) and the cutting result C 10 The system generates a target transmission scheme package based on the transmission rule list. The target transmission scheme package is defined as a path-based transmission scheme that can be directly invoked in subsequent economic evaluations. To encapsulate rule fragments, a weighted fusion function is introduced. Formula (11) is defined as follows: In the formula: R 11 w represents the weighted fusion result vector of the rule fragments; 11 (p d ) represents the p-th d The weights of each rule segment; p d : indicates the p-th d The index of a rule fragment; r 11 (p d ) represents the p-th d The original value of each rule fragment; P represents the number of rule fragments.

[0074] It is understandable that rule fragments are extracted from the list of propagation rules and denoted as r. 11 (p d ), R of formula (11) 11 The internal structure of the solution package is formed.

[0075] Finally, to manage version consistency, a stability check based on Lyapunov functions is introduced. Formula (12) is defined as follows: In the formula: V 12 The Lyapunov function value representing version deviation; ΔR 11 The change in the result of rule fragment fusion is a system parameter preset based on business risk preferences.

[0076] It is understandable that R from formula (11) 11 Calculate the change, V in formula (12) 12 Used to ensure a smooth version transition. R in formula (11) 11 As input to formula (12).

[0077] C3: Load the lifecycle parameter mirror onto the target transmission scheme package, and combine it with external pricing information and market transaction scenario data to perform cash flow forecast calculations under multiple scenarios, generating a multi-scenario economic evaluation set.

[0078] It should be noted that the input for this step is the target transmission scheme package, and within the same session, it reads only the lifecycle parameter mirror, the policy electricity price mirror, and the scenario data cache of the market transaction interface. Specifically, the assembly rule fragment set and path index of the target transmission scheme package are loaded into the evaluation channel. The allocation configuration of each path in each time segment is used as the prediction unit. At the same time, fixed field names such as construction input, operation and maintenance expenditure, residual value, and taxes in the lifecycle parameter mirror are called, and the rate version, time period classification, and tier boundary of the policy electricity price mirror, as well as the transaction price and transaction volume sequence in the scenario data cache are read. The processing flow first binds parameters to the prediction units. The parameter binding registers the allocation caliber in the assembly rule fragments and the cost items in the lifecycle parameter mirror. Then, the rate items in the policy electricity price mirror are mapped to the prediction units according to the time segment to form a combined item for cash flow calculation. During the parameter binding process, if it is found that a certain cost item does not have a corresponding allocation caliber in the assembly rule fragment, the system does not perform estimation, but directly registers the "unmapped" status on the combined item, and retains this status in the subsequent summary for write-back. After parameter binding is completed, the system performs cash flow forecasting calculations. Without altering the original item values, the cash flow forecasting calculations accumulate and merge cost and revenue items according to time segment order, forming a cash flow sequence by path and time segment. During account period merging, the system reads the account period anchor point of the session timeline, merges and registers items within the same account period, and writes the account period number and merging rule pointer into the merge record. Subsequently, the system generates a scenario set in the evaluation channel. The scenario set is composed of different rate versions of the policy electricity price mirror, different transaction sequence tables of the transaction data cache, and missing factor substitution strategies for assembly rule segments. Each scenario generates an independent cash flow summary item and records the scenario number and source pointer. If a scenario involves a missing factor substitution strategy, the system only records the "uncalcifiable" status and skips the combination without performing any numerical substitution. After completing the cash flow sequences and summary entries for each scenario, the system calculates and registers profitability indicators. In this embodiment, profitability indicators refer to the indexed representation of cumulative cash flow and time-series consolidation results. Specific field names include static recovery-related indicators, investment return-related indicators, and payment period performance-related indicators. Scenario number, path index, and function tag are bound to the indicators during registration. Through the above processing, the system generates a multi-scenario economic evaluation set, defined as "the set of economic results for path-based transmission schemes under different policy electricity prices and transaction conditions within the same session." Set entries include fixed field names such as scenario number, function tag, path index, cash flow sequence, payment period consolidation record, and profitability indicators. Version number and session number are registered at the head of the set.Ultimately, the output of this step is a multi-scenario economic assessment set, with the output field named "Multi-scenario Economic Assessment Set". This output is naturally input into the "Multi-scenario Economic Assessment Set" of the subsequent step D1 to generate the set of disturbance parameters and the disturbance processing of the allocation factor. At the same time, it provides a basic data source for sensitivity and resilience analysis at the cross-main step level, and retains the reference relationship between the target transmission scheme package and the policy electricity price mirror within the session.

[0079] Preferably, this step involves parameter binding and cash flow forecasting of the target transmission scheme package, and the construction of various policy and transaction scenarios to form a structured multi-scenario economic assessment set, providing a consistent data foundation oriented towards path and functional labels for subsequent disturbance and sensitivity analysis.

[0080] In an optional implementation, the method for obtaining pipeline hazard points in step S300 can also be based on uncertainty quantification using Monte Carlo simulation. That is, after the transmission rules are assembled, probability distribution models (such as normal distribution or empirical distribution) are constructed for key parameters such as external electricity price, market transaction price, energy load, and loss rate. Thousands of parameter combination scenarios are generated by random sampling, and each scenario is substituted into the cash flow forecasting model for parallel calculation. Finally, a multi-scenario economic evaluation set containing expected net present value, risk value, and confidence interval is formed to quantify the robustness of the cost diversion scheme under parameter fluctuations.

[0081] In another optional implementation, the method for obtaining pipeline hazard points in step S300 can also be through a historical data-driven approach based on machine learning. That is, collecting historical operating data of microgrids, market transaction records and settlement data as training sets, using cost aggregation unit boundaries, transmission rule factors, time period configurations, etc. as feature inputs, constructing LSTM or XGBoost models to learn the implicit mapping relationship between cost flow and economic indicators, and using the trained model to infer and predict different functional positioning path combinations, quickly generating an evaluation set covering multiple policies and operating conditions, thereby improving evaluation efficiency and model generalization.

[0082] In this embodiment of the application, step S400 involves performing parameter perturbation, sensitivity, and resilience analysis based on the multi-scenario economic evaluation set, identifying key risk factors, generating allocation suggestions, and then writing back and versioning the transmission rule parameters to generate an iteratively corrected cost diversion mechanism, including the following steps D1-D3: D1: Based on the multi-scenario economic evaluation set, generate a set of disturbance parameters and perform disturbance processing on the allocation factor to form a sensitive input set.

[0083] It should be noted that the input for this step is a multi-scenario economic evaluation set, and within the same session, the target transmission scheme package and functional positioning path set are read-only referenced as the configuration basis for the disturbance boundary. Specifically, the multi-scenario economic evaluation set is loaded as a session view in the evaluation channel. The session view is presented in segments according to functional labels, path indexes, and billing period numbers. Each segment contains fixed field names such as cash flow sequence, billing period consolidation record, and profitability indicators. After loading, the disturbance parameter set generation process is started. The disturbance parameter set is used to describe the parameter group that performs bounded disturbances on key configurable parameters without changing the original observation and assembly records. The parameter items at least cover fields such as electricity price item, transaction sequence, allocation caliber, capacity limit, loss range, billing period length, and discount strategy. The generation process first reads the rate version and time period classification consistent with the multi-scenario economic assessment set from the policy electricity price mirror, and constructs an external price disturbance subset at the version dimension. Then, it reads the assembly rule fragment set from the target transmission scheme package, and constructs a disturbance subset at the rule dimension, defining the allocation caliber and factor effective range. Next, it reads the switching indications and path sequences from the functional positioning path set, and constructs a disturbance subset at the structural dimension, defining the path switching points and continuous fragment lengths. For each subset, the system generates discrete disturbance points within the legal interval with a fixed step size, aligns the disturbance points with the session timeline, and writes them into a disturbance parameter set entry. This entry includes fixed field names such as parameter name, disturbance direction, disturbance amplitude, effective time segment, and source pointer. When a parameter has no legal interval or reference relationship in the session context, the system only registers the parameter's "undisturbable" state without making any estimations or replacements.

[0084] After the disturbance parameter set is generated, the system enters the allocation factor disturbance processing stage. This stage projects the disturbance parameters to the allocation factor level and forms a computable input slice. Specifically, the system reads three types of fields from the target transmission scheme package: loss allocation caliber, capacity allocation caliber, and time period allocation configuration, and performs segment-by-segment matching on the session timeline. For disturbance points belonging to the price category, a new shadow price entry is added and a parallel reference relationship is established with the original entry without modifying the original rate entry. The shadow price entry is only effective within the sensitivity assessment session. For disturbance points belonging to the transaction category, the system establishes an alternative transaction sequence in the transaction data cache and writes an alternative sequence identifier, which is then attached to the same time segment as the original transaction sequence. For disturbance points belonging to the assembly category, the system creates a temporary branch in the transmission rule list and performs interval replacement of the allocation caliber or factor effective range within the disturbance amplitude. The replacement action is strictly constrained by the boundaries and switching instructions of the assembly rule segment, and does not cross undefined path boundaries or merge non-existent segments. To avoid mutual interference between perturbations, the system adopts a session scheduling strategy that prioritizes single-factor perturbations and supplements them with two-factor perturbations. In each session segment, the input slice of single-factor perturbations is calculated first, and the generation of slices of two-factor perturbations is only initiated when it is necessary to identify joint effects. All slices are labeled with perturbation parameter identifiers, original entry pointers and calculation batch identifiers for subsequent result location and traceability.

[0085] After completing the perturbation processing of the allocation factor, the system merges the shadow price entries, alternative transaction sequences, and temporary branch rules for each time segment to form a set of slices to be evaluated. During the merging process, fields with the same name from different sources are prioritized and merged. The priority order is based on the source level of the transmission rule list and the priority label in the assembly log, and the covered fields are recorded in the coverage record for review. Finally, the system generates a sensitivity input set, defined as "a sliced ​​input set driven by the perturbation parameter set and aligned one-to-one with the multi-scenario economic assessment set". The set entries include fixed field names such as function label, path index, time segment, shadow price identifier, alternative transaction identifier, temporary rule branch identifier, and original entry pointer. Finally, the output of this step is the sensitivity input set, with the output field name "sensitivity input set". This output is naturally input into the "sensitivity input set" of the subsequent step D2 and serves as the sole data source for sensitivity and resilience analysis at the cross-main step level. At the same time, it retains a read-only reference relationship with the multi-scenario economic assessment set, target transmission scheme package, and functional positioning path set within the session for subsequent backtracking and comparison.

[0086] D2: Extract indicator sequences and path factor sequences from the aforementioned sensitivity input set, perform sensitivity and resilience analysis, and generate a list of key risk factors and allocation suggestions.

[0087] It should be noted that the input for this step is a sensitive input set, and within the same session, the profitability indicators and payment period merging records of the multi-scenario economic assessment set are read-only referenced as the basis for result alignment. Specifically, the system first performs a slice traversal of the sensitive input set, reading slice entries by function tag, path index, and time segment, and retrieves the profitability indicator sequence and cash flow sequence consistent with the slice entry from the multi-scenario economic assessment set, aligning them one-to-one in session number and time segment. After alignment, the system enters the extraction process of indicator sequence and path factor sequence. The indicator sequence refers to the profitability indicator time series corresponding to the original scenario under disturbance conditions and the indicator sequence after payment period merging. The path factor sequence refers to the value and switching trajectory of factors such as allocation caliber, loss range, capacity limit, and time period configuration on the time segment under disturbance conditions. The extraction process reads the shadow price identifier, substitute transaction identifier, and temporary rule branch identifier for each slice entry, and constructs the perturbation indicator sequence and perturbation path factor sequence within the session. When a slice entry lacks any necessary identifier or an anomaly occurs in the signature, the entry is marked as an uncalculated entry and excluded from this analysis batch. At the same time, the reason, source, and suggested re-sampling window are recorded in the anomaly registration.

[0088] After the indicator sequence and path factor sequence are extracted, the system performs sensitivity and elasticity analysis on each analysis batch. Sensitivity analysis is used to characterize the magnitude and direction of changes in profitability indicators and cash flow sequences when a single disturbance is applied to a parameter; elasticity analysis is used to characterize the distribution characteristics of stable and unstable regions of indicator response in different intensity ranges when the disturbance acts on the parameter proportionally. The processing flow first performs sensitivity calculations on a single-factor slice basis, correspondingly reading the disturbance magnitude, time segment, and affected factor items in the path factor sequence of the disturbance parameter. Within the session, the difference and segment change points between the indicator sequences before and after the disturbance are calculated, and the segment change points are associated with the path factor switching indication to identify response changes caused by rule switching. Subsequently, on the paths and time segments where joint effects need to be identified, two-factor slices are read for comparison. The indicator difference in the two-factor slice is compared with the difference between the two single-factor slices. If a relative amplification or relative cancellation phenomenon occurs, it is recorded as an interaction marker, and the participating factors, time segments, and strength descriptions are written in the interaction record. For elasticity analysis, the system generates multi-level disturbance points within the legal range of each disturbance parameter (such as the range of electricity price change from -10% to +10%) according to a preset step size, and performs segmented stability tests on the index sequence of each level of disturbance point. When a phased stability or abrupt change of the index sequence occurs near a certain level of disturbance point, the system writes the start and end boundaries of the stable or unstable segment, the corresponding factors and paths into the elasticity record, forming an intervalized characterization of disturbances of different intensities.

[0089] After completing sensitivity and resilience analysis, the system integrates the results of single-factor and two-factor batches to construct a key factor candidate set for each functional label and path index. The key factor candidate set originates from factors with significant sensitivity differences, unstable resilience segments, and interaction markers. For the factor items in the candidate set, the system sorts them by source level, path priority, and session coverage, generating a key risk factor list after sorting. The key risk factor list is defined as "a list of factors that have a significant impact on profitability indicators and cash flow sequences and have interpretable paths under the current session and the aforementioned multi-scenario conditions." This list includes fixed field names such as functional label, path index, factor category, trigger fragment, interaction marker, and sorting position. Based on the key risk factor list, the system generates allocation suggestions according to the switching instructions and assembly rule fragment boundaries of the functional positioning path set. These allocation suggestions include parameter adjustment suggestions for loss allocation caliber, capacity allocation caliber, and time period allocation configuration, path weighting suggestions, and switching boundary refinement suggestions. Each suggestion entry records the suggestion direction, suggestion scope, related factors, and corresponding evidence pointers, without rewriting or estimating the original data. Ultimately, the output of this step is a list of key risk factors and allocation recommendations, with the output field named "List of Key Risk Factors and Allocation Recommendations". This output is naturally input into the "List of Key Risk Factors and Allocation Recommendations" in the subsequent step D3, and serves as the direct basis for rule parameter write-back and version archiving at the cross-main step level. At the same time, it retains the reference relationship with the sensitivity input set and the multi-scenario economic assessment set within the session to support subsequent backtracking and auditing.

[0090] D3: Based on the list of key risk factors and the allocation suggestions, write back and archive the transmission rule parameters to generate an iteratively revised cost diversion mechanism.

[0091] It should be noted that the input for this step is a list of key risk factors and allocation recommendations, and within the same session, the target transmission scheme package and transmission rule list are read-only references as the objects of parameter write-back. Specifically, the system first establishes a parameter write-back context in the write-back channel, groups the list of key risk factors by function tag and path index, and aligns them with the recommendation direction and scope in the allocation recommendations one by one; for each aligned entry, the system locates the corresponding assembly rule fragment and factor configuration item in the transmission rule list, locates the action read node pair, direction mark and effective fragment boundary, and checks the version number and source signature of the configuration item in the current session. After successful location, the system executes a parameter write-back strategy based on the content of the allocation suggestion. The parameter write-back strategy follows the basic principles of "not overwriting the original value, storing versions in parallel, and explicitly switching references." Specifically, the new parameter is written into the assembly rule fragment in the form of a new version, and the version number, source signature, and write-back reason are registered in the fragment header. The old and new version entries are retained in the index of the transmission rule list, and the reference switching point and switching conditions are recorded in the index. For suggestions that require path weighting, the system creates a temporary branch of the weight table on the functional location path set, and records the weighted path, reduced path, and effective time fragment in the weight table branch. The temporary branch is turned into a regular branch after the new version scheme takes effect. If it fails the audit, it is deleted at the end of the session.

[0092] After completing the parameter write-back, the system performs consistency checks and impact scope marking. Consistency checks examine the boundary consistency and source compliance of the new version parameters between the assembly rule fragment and the factor effective fragment set. If boundary inconsistencies or source signature mismatches occur, the write-back entry is marked as "pending review" and will not be included in the publicly released version. Impact scope marking marks the time segments, node pairs, and switching points involved in the new version parameters on the functional positioning path set, providing location information for subsequent re-evaluation and rolling release. Immediately afterwards, the system initiates the version archiving process. Version archiving uses the session number as an index to archive the new version parameters and temporary branches generated by this write-back. The archived entries record the version number, write-back reason, associated key factors, associated allocation suggestions, approval status, and audit pointer. When the archiving action involves the synchronous reference of the policy electricity price mirror and transaction data cache, the system writes the mirror version reference in the archived record without copying the original mirror content, maintaining a read-only association. After the version is archived, the system outputs the iteratively corrected cost diversion mechanism at the mechanism layer. The iteratively corrected cost diversion mechanism is defined as "a set of executable cost diversion schemes consisting of new version propagation rules, updated path weights and archived version indexes". The set header registers the session number, release time and source signature. The set body contains the reference relationship and switching conditions between the old and new versions, and provides a read-only retrieval interface to the outside world.

[0093] The output product is the iteratively corrected cost channeling mechanism, with the output field named "Iteratively Corrected Cost Channeling Mechanism". This output is written back to the transmission rule assembly parameter position of C2 and the economic probe configuration position of B1 at the cross-main step level, which are used to drive the generation of a new round of assembly rule fragments and the updating of the probe acquisition strategy, respectively. At the same time, this output forms a closed loop reference with the multi-scenario economic evaluation set and the sensitivity input set within the session, providing a structured basis for the re-evaluation and continuous auditing of the next business cycle.

[0094] Understandably, iterations are automatically triggered by data analysis results, and most processes are executed automatically. However, an automatic brake is set at critical stages through the "pending review" status, ensuring that any potentially problematic fixes will not take effect automatically. This allows the final decision-making power to be handed over to humans, achieving "automated execution, controlled release." Furthermore, the version management strategy ensures stability through read-only mirroring, parallel storage, explicit switching, and session isolation, while ensuring ultimate auditability through session numbering, full metadata recording, and source signatures. This makes the entire cost-sharing mechanism's evolution not only highly automated but also secure, controllable, transparent, and trustworthy.

[0095] Preferably, this step involves writing back parameters and archiving versions of the list of key risk factors and allocation recommendations to form an iteratively revised cost transfer mechanism with version traceability and switching conditions. This enables continuous updating and traceable management of the cost transmission scheme without rewriting the original observations and historical rules.

[0096] In an alternative implementation, the method for obtaining pipeline hazard points in step S400 can also be achieved through an adaptive parameter optimization method based on reinforcement learning. This method uses the transmission rule parameter space as the action space of the agent, the profitability index of the multi-scenario economic evaluation set as the reward signal, and constructs a Markov decision process for microgrid cost mitigation. Through Q-learning or policy gradient algorithms, the agent is allowed to repeatedly try and fail in a simulated environment to automatically learn the optimal parameter adjustment strategy. When the reward converges or a preset threshold is triggered, the learned strategy is converted into a new version of the transmission rule and automatically archived.

[0097] In another optional implementation, the method of obtaining pipeline hazard points in step S400 can also be based on a multi-party collaborative iterative approach using federated learning. That is, when the microgrid involves multiple independent operating entities, each entity performs sensitivity analysis and generates allocation suggestions locally based on its own multi-scenario evaluation set. Only the encrypted parameter gradient or suggestion features are uploaded to the aggregation server. The server integrates the opinions of multiple parties through a federated averaging algorithm to generate global rule updates. After each entity downloads the update, it makes fine adjustments based on local business constraints and completes version archiving. The original data does not leave the domain throughout the process, forming a distributed iterative correction mechanism under privacy protection.

[0098] In summary, the beneficial effects of this microgrid cost mitigation mechanism design and evaluation method are as follows: In the data acquisition and alignment stage, the use of unified session numbering and precise time anchor registration ensures the full traceability and temporal consistency of multi-source heterogeneous data across billing periods; in the rule assembly stage, dynamic filtering based on functional positioning and intelligent pruning of conflict priorities enables the structured organization of transmission rules within time segments and stable integration with economic evaluation; in the evaluation iteration stage, key risk factors are identified based on parameter disturbance and sensitivity analysis, driving closed-loop write-back and version management of rule parameters, forming an auditable and evolvable long-term mechanism. These three aspects are interconnected, spanning the entire lifecycle of "acquisition-assembly-evaluation-optimization," significantly improving the transparency, adaptability, and continuous optimization capabilities of microgrid cost mitigation.

[0099] Example 3 illustrates a schematic scheme for the design and evaluation method of a microgrid cost mitigation mechanism. It should be noted that the technical solution of this microgrid cost mitigation mechanism design and evaluation system belongs to the same concept as the technical solution of the aforementioned microgrid cost mitigation mechanism design and evaluation method. Details not described in detail in this example can be found in the description of the technical solution of the aforementioned microgrid cost mitigation mechanism design and evaluation method.

[0100] This embodiment also provides a microgrid cost mitigation mechanism design and evaluation system, including: The first generation module is used to unify data scope and divide cost unit boundaries based on the functional definition information and entity relationship mapping information of the microgrid, and generate a structured cost collection unit delimitation structure. The module is used to collect and verify traceable data based on market transaction data, external pricing information and energy metering and settlement data, and to construct a cost transmission diagram that reflects the flow path and factors of costs between units according to the cost collection unit boundary structure. The prediction module is used to assemble cost transmission rules based on the cost collection unit boundary structure and the cost transmission diagram, and load life cycle parameters to predict cash flow under multiple scenarios, generating a multi-scenario economic evaluation set. The second generation module is used to perform parameter perturbation, sensitivity and elasticity analysis based on the multi-scenario economic evaluation set, identify key risk factors and generate allocation suggestions, and then write back and version control the transmission rule parameters to generate an iteratively corrected cost diversion mechanism.

[0101] This embodiment also provides an electronic device applicable to the design and evaluation of microgrid cost mitigation mechanisms, comprising: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the microgrid cost mitigation mechanism design and evaluation method proposed in the above embodiment.

[0102] This embodiment also provides a storage medium on which a computer program is stored. When the program is executed by a processor, it implements the design and evaluation method for the microgrid cost mitigation mechanism proposed in the above embodiments.

[0103] The storage medium proposed in this embodiment and the design and evaluation method for realizing the cost mitigation mechanism of microgrids proposed in the above embodiments belong to the same inventive concept. Technical details not described in detail in this embodiment can be found in the above embodiments, and this embodiment has the same beneficial effects as the above embodiments.

[0104] Based on the above description of the implementation methods, those skilled in the art can clearly understand that the present invention can be implemented using software and necessary general-purpose hardware, and of course, it can also be implemented using hardware. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as a computer floppy disk, read-only memory (ROM), random access memory (RAM), flash memory, hard disk, or optical disk, etc., including several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods of the various embodiments of the present invention.

[0105] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A design and evaluation method for a microgrid cost mitigation mechanism, characterized in that: include, Based on the functional definition information and entity relationship mapping information of the microgrid, data caliber is unified and cost unit boundaries are divided to generate a structured cost collection unit delimitation structure. Based on market transaction data, external pricing information, and energy metering and settlement data, traceable data collection and verification are carried out, and a cost transmission diagram reflecting the flow path and factors of costs between units is constructed according to the cost collection unit boundary structure. Based on the cost collection unit delimitation structure and the cost transmission diagram, cost transmission rules are assembled, and life cycle parameters are loaded to predict cash flow under multiple scenarios, generating a multi-scenario economic evaluation set. Based on the multi-scenario economic assessment set, parameter perturbation, sensitivity and elasticity analysis are performed to identify key risk factors and generate allocation suggestions. Then, the transmission rule parameters are written back and versioned to generate an iteratively corrected cost diversion mechanism.

2. The microgrid cost mitigation mechanism design and evaluation method as described in claim 1, characterized in that: The traceable data collection and verification includes, Read-only economic probes are deployed at the metering, settlement, and market trading ends to collect key data fields; Bind a unique session number to all collected data entries and attach a source signature and time anchor; The collected data is packaged into a traceable cost-benefit observation stream.

3. The microgrid cost mitigation mechanism design and evaluation method as described in claim 2, characterized in that: Constructing the cost transmission diagram includes, Using the unit identifiers in the cost aggregation unit delimitation structure as nodes and the cost flow relationships in the cost-benefit observation flow as edges, node-edge mapping is performed. By linking the loss allocation factor, capacity allocation factor, and time period allocation factor to the corresponding paths, a cost transmission graph containing a set of nodes, a set of edges, and a factor-path association table is generated.

4. The microgrid cost mitigation mechanism design and evaluation method as described in claim 1, characterized in that: The assembly of the cost transmission rules includes, Extract the loss allocation factor, capacity allocation factor, and time period allocation factor that are bound to the functional path from the cost transmission diagram. The factors are merged and associated according to nodes and directions to generate assembly rule fragments; When rule conflicts exist within the same time segment, they are pruned according to preset priorities, retaining the higher-priority rules to form a list of propagation rules.

5. The microgrid cost mitigation mechanism design and evaluation method as described in claim 1, characterized in that: The parameters are subjected to perturbation, sensitivity, and elasticity analysis. include, Based on the aforementioned multi-scenario economic assessment set, a set of disturbance parameters for electricity price, allocation method and path structure is generated; The perturbation parameters are projected onto the allocation factor level to form an input slice for sensitivity analysis; By calculating the differences and change points of economic indicators before and after the disturbance, the sensitivity and elasticity of single parameters and joint parameters are analyzed.

6. The microgrid cost mitigation mechanism design and evaluation method as described in claim 1, characterized in that: The generated iteratively corrected cost mitigation mechanism includes, The allocation recommendations generated based on key risk factors will be written back to the corresponding parameters in the list of transmission rules in the form of a new version; After writing back the rules, a consistency check and version archiving are performed to form a set of executable cost mitigation solutions that include version reference relationships and switching conditions.

7. A method for designing and evaluating a microgrid cost mitigation mechanism as described in any one of claims 1-6, characterized in that: The external pricing information refers to read-only mirror data formed after the structured rate schemes, subsidy items, and tiered pricing boundaries from outside the microgrid are version-fixed.

8. A microgrid cost mitigation mechanism design and evaluation system, using the method described in any one of claims 1-7, characterized in that, include: The first generation module is used to unify data scope and divide cost unit boundaries based on the functional definition information and entity relationship mapping information of the microgrid, and generate a structured cost collection unit delimitation structure. The module is used to collect and verify traceable data based on market transaction data, external pricing information and energy metering and settlement data, and to construct a cost transmission diagram that reflects the flow path and factors of costs between units according to the cost collection unit boundary structure. The prediction module is used to assemble cost transmission rules based on the cost collection unit boundary structure and the cost transmission diagram, and load life cycle parameters to predict cash flow under multiple scenarios, generating a multi-scenario economic evaluation set. The second generation module is used to perform parameter perturbation, sensitivity and elasticity analysis based on the multi-scenario economic evaluation set, identify key risk factors and generate allocation suggestions, and then write back and version control the transmission rule parameters to generate an iteratively corrected cost diversion mechanism.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 7.

Images