Green power direct connection project full life cycle cost fine calculation method, system, device and medium

Abstract

The invention relates to the technical field of enterprise management and cost accounting, in particular to a green electricity direct connection project full life cycle cost fine measurement and calculation method, system and device and a medium, and the method comprises the steps: constructing a dynamic cost conduction path network in combination with predefined allocation rules and trigger conditions, carrying out cost collection along the path network, and carrying out cost calculation. Generating project-level cost summary data; and the project-level cost summary data is bound with external scene parameters to generate a plurality of measurement and calculation scenes, full-life-cycle economy accounting is performed on each scene to obtain an economy index set, multi-objective optimization analysis is performed on the economy index set, and decision support information is synthesized to generate a structured decision packet. The method has the beneficial effects that a closed-loop link from heterogeneous data cleaning and dynamic cost conduction to multi-target scene comparison and selection is constructed, so that the caliber unification, process transparency and decision-making robustness of green electricity direct connection project full life cycle cost measurement and calculation are realized.

Description

Technical Field

[0001] This invention relates to the field of enterprise management and cost accounting technology, and in particular to a method, system, equipment and medium for precise calculation of the full life cycle cost of green electricity direct connection projects. Background Technology

[0002] With the deepening of market-oriented reforms in green electricity trading, green electricity direct-connection projects involve multiple stages throughout their lifecycle, including investment, construction, grid connection, operation, and decommissioning. Cost calculation for these projects requires integrating heterogeneous data from multiple stakeholders, including owners, electricity users, the grid, and energy storage operators. Current industry practice commonly employs decentralized contract management systems, grid connection metering systems, equipment operation and maintenance platforms, and financial accounting systems to aggregate data at different stages, supplemented by manually set scenario parameters and coarse-grained discounting models to complete project economic comparisons.

[0003] However, the following problems exist in actual use: First, the data standards of various entities are not unified, and there is a lack of a global index. This leads to each system operating independently, such as contract data, equipment maintenance records, and financial invoices. This not only makes cross-system consistency verification impossible but also makes it difficult to trace the source of problems. Second, the cost transmission and allocation process is not transparent, and the allocation rules rely solely on static contract terms without trigger condition configuration. When contract changes, metering anomalies, or maintenance work order updates occur, the aggregation path cannot be dynamically adjusted, causing interruptions in the cost accounting chain. Furthermore, the boundary binding between scenario parameters and project calculations is unstable. When policies or commercial conditions such as electricity pricing mechanisms and capacity configurations change drastically, the mapping relationship between parameter bits and data fields is prone to failure, affecting the reliability of scenario comparison. Even more challenging is the lack of robust methods for handling missing, fluctuating, or inter-period data, leading to distorted accounting results. Moreover, during the decision-making stage, multiple objectives are often weighed against subjective weighting. These problems make it difficult for the entire calculation chain to meet the sophisticated management needs of today's complex market. Summary of the Invention

[0004] To solve the above-mentioned technical problems, the present invention provides the following technical solution: In the first aspect, the present invention provides a method for precise calculation of the full life cycle cost of green electricity direct connection projects, including subject identification and timeline standardization based on basic project information and investment and construction mode, as well as the deployment and index binding of multi-source data acquisition probes to generate a probe primary key index structure; Based on the probe's primary key index structure, multi-source data on contracts, equipment, and finance are collected, and the collected data is preprocessed to generate standard data mirrors and abnormal event records. Data normalization is performed based on standard data mirroring to identify cost occurrence points. Combined with predefined allocation rules and triggering conditions, a dynamic cost transmission path network is constructed. Costs are collected along the path network to generate project-level cost summary data. Project-level cost aggregate data is bound to external scenario parameters to generate multiple calculation scenarios. Full life-cycle economic accounting is performed on each scenario to obtain a set of economic indicators. Multi-objective optimization analysis is performed on the economic indicator set, and decision support information is synthesized to generate a structured decision package.

[0005] As a preferred embodiment of the method for precise calculation of the full life-cycle cost of green electricity direct connection projects according to the present invention, the deployment of multi-source data acquisition probes includes: Deploy contract-transaction probes for collecting contract execution and transaction settlement data, equipment-maintenance probes for collecting equipment operation and maintenance data, and financial-invoice probes for collecting financial voucher and invoice data.

[0006] As a preferred embodiment of the method for precise calculation of the full life-cycle cost of green electricity direct connection projects according to the present invention, the method includes: identifying cost incurrence points, including... Identify settlement, default, or price adjustment entries in the contract execution record as contract-related cost incurrence points; Identify billing entries related to capacity, electricity consumption, or ancillary services in grid-connected metering records as transaction cost incurrence points; Replacement, repair, or decommissioning / reactivation entries in equipment operation and maintenance records are identified as points where operation and maintenance costs occur.

[0007] As a preferred embodiment of the method for precise calculation of the full life cycle cost of green electricity direct connection projects in this invention, the dynamic cost transmission path network is a directed graph structure; In this directed graph structure, the nodes are defined by the main body, cost unit, and time point. In this context, the edges of the directed graph structure are defined by the allocation rules and triggering conditions, which are used to represent the direction of cost transmission and allocation.

[0008] As a preferred embodiment of the method for precise calculation of the full life-cycle cost of green electricity direct connection projects according to the present invention, the method further includes, when constructing a dynamic cost transmission path network, the following: Read the generated exception event logs; Set segment nodes on the path segments covered by the abnormal event log; Mark soft boundary information and suggested processing paths in the edge attributes associated with segment nodes.

[0009] As a preferred embodiment of the method for precise calculation of the full life-cycle cost of green electricity direct connection projects according to the present invention, the method includes: performing multi-objective optimization analysis on the set of economic indicators, including, Robust sorting is performed on the set of economic indicators to reduce the impact of missing data and jumps on the sorting results; The ranking results are subjected to Pareto screening to identify non-dominated solution sets as candidate target scenarios.

[0010] As a preferred embodiment of the method for precise calculation of the full life-cycle cost of green electricity direct connection projects according to the present invention, the method includes: collecting multi-source data on contracts, equipment, and finance based on the probe primary key index structure, and preprocessing the collected data, including... Based on the acquisition channels and triggering conditions defined in the probe primary key index structure, contract execution records, equipment operation and maintenance records, and financial invoice records are acquired in parallel. The collected multi-source data undergoes source signature verification and clock alignment processing. Perform missing test labeling and abnormal segment identification on the aligned and checked data; Based on the identification results, a standard data image containing cleaned data is generated, as well as an anomaly event log that records anomaly information and processing suggestions.

[0011] Secondly, the present invention provides a system for precise calculation of the full life cycle cost of green electricity direct connection projects, including: a generation module, used to perform main body identification and time axis standardization based on basic project information and investment and construction mode, as well as the deployment and index binding of multi-source data acquisition probes, and generate a probe primary key index structure; The preprocessing module is used to collect multi-source data on contracts, equipment, and finance based on the probe's primary key index structure, and to preprocess the collected data to generate standard data images and abnormal event records. The module is used to perform data normalization processing based on standard data mirroring, identify cost occurrence points, and build a dynamic cost transmission path network by combining predefined allocation rules and triggering conditions. Costs are collected along the path network to generate project-level cost summary data. The synthesis module is used to bind project-level cost summary data with external scenario parameters to generate multiple calculation scenarios. For each scenario, full life cycle economic accounting is performed to obtain a set of economic indicators. Multi-objective optimization analysis is performed on the set of economic indicators, and decision support information is synthesized to generate a structured decision package.

[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, wherein the computer program, 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: A closed-loop system for full lifecycle cost calculation is constructed through the synergistic effect of three major links: First, at the data acquisition end, relying on a three-dimensional index and robust noise reduction mechanism, only missing data is tagged without modifying the original data, and abnormal behavior is precipitated into a traceable event queue, providing stable and reliable data input for subsequent processing. Second, at the cost aggregation end, through a dynamic cost transmission graph, the relationship between fields, allocation rules, and triggering conditions is solidified into a topological structure of nodes and edges, achieving transparency and version replayability of multi-entity, multi-stage cost allocation paths, ensuring accurate aggregation in scenarios with differences in caliber and weak correlations. Finally, at the scenario comparison end, through dynamic parameter binding and robust sorting-Pareto screening mechanisms, agile responses are made to changes in policies, contracts, and capacity schemes, generating a visual decision package bound to the interpretation view and writing back to the dictionary for iterative optimization. These three links are interconnected, forming a complete closed loop from data cleaning and cost transmission to intelligent decision-making, significantly improving the multi-entity collaborative efficiency, cost accounting accuracy, and standardized decision-making capabilities of green electricity direct connection projects in complex market environments. 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 method for detailed calculation of the full life cycle cost of green electricity direct connection projects. 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 detailed calculation of the full life cycle cost of a green electricity direct connection project, including: S100: Based on the project's basic information and investment and construction model, standardize the main body identification and timeline, and deploy and bind multi-source data acquisition probes to generate a probe primary key index structure. S200: Based on the probe's primary key index structure, collect multi-source data on contracts, equipment, and finance, and preprocess the collected data to generate standard data mirrors and abnormal event records; S300: Based on standard data mirroring, it performs data normalization processing, identifies cost occurrence points, and combines predefined allocation rules and triggering conditions to construct a dynamic cost transmission path network. Costs are collected along the path network to generate project-level cost summary data. S400: Binds project-level cost aggregate data with external scenario parameters to generate multiple calculation scenarios. Performs full lifecycle economic accounting on each scenario to obtain a set of economic indicators. Performs multi-objective optimization analysis on the economic indicator set and synthesizes decision support information to generate a structured decision package. It should be noted that this method, through steps S100-S400, breaks down heterogeneous data barriers between multiple entities and systems by constructing a unified probe primary key index structure. This achieves standardization of the entire process from project information initialization to cost aggregation and robust anomaly handling, improving data quality and traceability. Based on the construction of a dynamic cost transmission path network, static allocation rules are transformed into a flexible mechanism that can be automatically triggered by contract changes, policy adjustments, and maintenance events, making the cost aggregation process transparent and auditable. Through dynamic binding of scenario parameters and project-level cost data and multi-objective optimization analysis, it supports agile scenario comparison under changing commercial conditions such as electricity pricing mechanisms and capacity configuration. Combining robust sorting and Pareto screening effectively quantifies the multi-objective trade-offs, avoiding subjective judgment bias.

[0019] Example 2, refer to Figure 1 As an embodiment of the present invention, based on the above embodiment, a method for online health monitoring of high-temperature pipelines is provided.

[0020] In this embodiment of the application, step S100 involves standardizing the main entity identifier and timeline based on the project's basic information and investment and construction model, as well as deploying and indexing multi-source data acquisition probes to generate a probe primary key index structure, including the following steps A1-A4: Understandably, the basic project information is a collection of general data generated at each stage of project initiation, grid connection, transaction, operation, and decommissioning, including at least the project name, geographical location, installed capacity, grid connection voltage level, transaction category, contract list, and grid connection approval information; the investment and construction model is a collection of combinations of engineering construction and funding organization, including at least self-built and self-used, contract energy management, construction and operation on behalf of others, equity cooperation and leasing, and summaries of their settlement terms.

[0021] A1: Uniquely identify and encode the responsible entities involved in the project to generate a primary key.

[0022] Understandably, the responsible entity refers to various independent entities involved in the incurrence, bearing, or settlement of costs in green electricity direct connection projects, such as investors, electricity users, grid companies, and operation and maintenance service providers. The unique code refers to assigning each entity a number or symbol identifier that remains unchanged and is not repeated throughout the entire project lifecycle.

[0023] A2: Determine the start and end times of each stage in the project's entire lifecycle, as well as the key event time points, to form a hierarchical time anchor structure.

[0024] It should be noted that after receiving the project acceptance file, the system breaks it down according to the dimensions of subject, stage, and data source. The subject dimension is used to distinguish between various independent but settlement-related responsible entities, including the owner, electricity user, grid connection and trading, energy storage and flexibility resources, compliance and carbon green certificates, and decommissioning management. The stage dimension is used to identify the time span and event sequence of each stage of investment, construction, grid connection, operation, maintenance, and decommissioning. The data source dimension is used to identify the source paths from contract texts, equipment ledgers, grid connection metering, trading platforms, and financial documents. Based on this, the system triggers the subject primary key binding process. The subject primary key is a fixed set of fields that uniquely identifies each responsible entity, including fields such as subject code, subject type, legal contact person, and contract association number. The primary key binding process checks the consistency between the subject ledger and the contract list, and triggers a manual confirmation node when there are entities with the same name or multiple contracts bound, in order to eliminate the problems of duplicate names, aliases, and cross-contract duplication. Furthermore, the system creates time anchors for each subject and stage event. A time anchor is a standardized time mark that can be referenced throughout the entire life cycle and is used to align the data collection and settlement sequence. The time anchor registration process writes the project initiation time, grid connection time, contract effective time, settlement cycle start and end time, maintenance event time, and decommissioning time into a searchable time index table, and forms a one-to-many hierarchical structure with stage start and end times and event times. After the primary key and time anchor are registered, the system proceeds to dimension and caliber processing. Dimension is a set of unit mappings used to express numerical measurement, and caliber is a set of caliber mappings used to describe the range of data and the boundary of calculation. The unified dimension and caliber dictionary is a standard dictionary used to resolve differences in units and calibers and to constrain the fields entered into the database when data is coordinated from multiple sources and at multiple stages. The system reads the dimensions of grid-connected metering, the dimensions of equipment ledgers, the subject caliber of financial vouchers, the settlement caliber of contract lists, and the quotation caliber of trading platforms through the dictionary generator. It uses field name mapping, field alias mapping, and priority determination rules to complete the unification and output a unified dimension / caliber dictionary that can be called later. Simultaneously, the system extracts strategy configurations that affect subsequent calculations from the project's basic information and investment and construction model to form a scenario parameter set. The scenario parameter set is a set of parameters called by the scenario calculation, which includes at least the direct power purchase price mechanism, green certificate price range, carbon price range, power purchase and sale contract structure, energy storage capacity configuration scheme, tax and subsidy parameters, and fields related to discounting and capital costs. At this stage, the scenario parameter set only completes initial registration and does not proceed with calculations.

[0025] After the above processing, this step, following the completion of primary key binding and time anchor registration, outputs a subject-stage mapping table, a unified dictionary of units / calibers, and a scenario parameter set. The subject-stage mapping table records the correspondence between subjects and stages, the stage time range, and event indexes. The unified dictionary of units / calibers records the units, caliber rules, and alias mappings for each field. The scenario parameter set records the initial parameter values ​​used for subsequent scenario matrix assembly. The scenario parameter set output in this section is called by subsequent step D4 when crossing main steps. It is important to emphasize that during the above operation, the system sets retry and rollback strategies for the primary key binding and time anchor registration. When inconsistencies or missing items occur, the corresponding fields are marked as pending verification and recorded in the exception registration table. The exception registration table is used as the verification basis in the subsequent data collection stage and does not affect the reference of the output fields in this step in subsequent steps.

[0026] Preferably, the hierarchical time anchor structure formed in this step provides a benchmark for clock alignment of subsequent data acquisition, ensuring the consistency of data across systems and sources in the time dimension, making time series analysis of costs and accurate mapping of cash flow possible.

[0027] A3: Based on the primary key and time anchor structure, deploy multiple data collection probes for collecting contract transaction data, equipment operation and maintenance data, and financial invoice data.

[0028] Understandably, a probe is essentially a software-defined, business-oriented data monitoring and acquisition unit. Specifically, it consists of a structured set of data acquisition configuration rules that are dynamically generated and maintained by the system. These rules are defined through a probe configuration table and parsed and executed by the system's data acquisition engine to automatically acquire standardized data from specified business systems (contracts, equipment, finance) according to preset primary keys, frequencies, and trigger conditions.

[0029] It should be noted that during this step, the system uses the output subject-stage mapping table as the primary key and stage reference for probe deployment, and uses the output unified dictionary of dimensions / calibers as the source of field constraints and unit conversion rules. It completes the deployment registration of three types of probes around four high-frequency data paths: contract execution, grid connection metering, operation and maintenance events, and invoice entries. Specifically, the contract-transaction probe is a data acquisition configuration for contract performance and transaction settlement scenarios. The probe objects are the power purchase and sale contract text, contract execution records, and transaction platform settlement data. The probe primary key consists of the contract number, settlement cycle, and subject code. The sampling interval refers to the data extraction frequency within each settlement cycle (e.g., monthly / quarterly or triggered during price adjustments). The acquisition channels are contract terms fields, performance status fields, and quotation and settlement fields. Triggering conditions include contract status changes, price adjustments, and the effectiveness of supplementary agreements. The probe deployment registration writes the above primary key, interval, channels, and triggering conditions into the probe configuration table and aligns them with the stage time range in the subject-stage mapping table. The Equipment-Maintenance Probe is a data acquisition configuration for equipment operation and maintenance activities. The probe objects include equipment tag number, operating condition, maintenance activity nodes, and stop / resume records. The probe primary key consists of the equipment tag number, stage event number, and subject code. The sampling interval is configured based on the frequency of operating data and maintenance activities (e.g., added every 15 minutes or when an alarm is triggered). Acquisition channels include power, energy, status, alarms, maintenance work orders, and repair records. Trigger conditions are status changes, alarm triggers, and maintenance work order changes. After deployment and registration, the Equipment-Maintenance Probe transmits data back within a time window aligned with the stage event. The Finance-Invoice Probe is a data acquisition configuration for financial accounts and invoice images. The probe objects include accounting vouchers, invoice images, and account images. The probe primary key consists of the invoice number, accounting date, and subject code. The sampling interval is configured according to the accounting period (e.g., by accounting month or triggered when an invoice is entered). Acquisition channels include accounting amount, tax rate, account number, and image pointer. Trigger conditions are invoice entry, reversal, and adjustment. Furthermore, the system constrains the fields of the three types of probes based on a unified dictionary of dimensions / calibers, unifying the settlement caliber in contract terms, the quotation caliber in the trading platform, the units in grid connection metering and equipment ledgers, and the account caliber in financial vouchers, thus avoiding ambiguity caused by different meanings or inconsistent units for fields with the same name. During the deployment and registration process, the system generates a probe configuration number for each probe configuration and records the mapping path between it and the subject-stage mapping table, forming a traceable configuration relationship. When overlapping collections across subjects, stages, or sources are identified, the system prioritizes the use of a finer-grained probe channel and records another path as a supplementary channel for subsequent consistency verification.After completing the above processing, the system summarizes the configuration entries of the contract-transaction probe, equipment-maintenance probe, and finance-invoice probe. Combining the subject and the time range of the stage, a probe deployment plan is prepared. The probe deployment plan is a list of data channels for collecting data on a specific primary key within a specific stage window, including sampling interval, triggering conditions, return format, and exception recording strategy.

[0030] It is important to emphasize that the probe deployment plan uses a unified dictionary of dimensions / calibers as a constraint reference when it is generated, and the semantics of the fields and the unit definitions will not be changed in the subsequent collection phase. At the same time, the return format in the probe deployment plan declares the required basic fields and optional fields. When optional fields are missing, the system only marks the missing fields without making an estimate or inference, and uses them as an anomaly clues in the subsequent consistency verification.

[0031] A4: Bind the contract number, equipment tag number, and invoice number to the corresponding probes using a three-dimensional index to generate the probe primary key index structure.

[0032] It should be noted that after receiving the output probe deployment plan, the three-dimensional index binding process is entered. The index key for the contract dimension is the contract number, the index key for the equipment dimension is the equipment tag number, and the index key for the invoice dimension is the invoice number. The three-dimensional index binding is the process of combining and referencing the primary key fields of the three types of probes according to the above three dimensions, aiming to form a unified primary key structure that can support parallel acquisition and playback query from multiple sources.

[0033] Specifically, the system first standardizes the contract number and version label for contract-transaction probe entries. The version label records the contract signing version, supplementary agreement version, and change record sequence number, and establishes a mapping from contract number to entity code, settlement cycle, and stage time range in the index. Then, it standardizes the device tag number for equipment-maintenance probe entries. The device tag number standardization process associates the device tag number with the device level, device function, and installation location, and generates a mapping from device tag number to entity code, stage event number, and time window. Finally, it standardizes the invoice number for finance-invoice probe entries. The invoice number standardization process associates the invoice number with the accounting date, accounting period, and account mirroring, and generates a mapping from invoice number to entity code and accounting period. After standardization across the three dimensions, the system performs index merging, cross-referencing the contract number, equipment tag number, and invoice number using subject coding and stage time to generate a cross-table for the three-dimensional index. For scenarios where contract terms directly relate to equipment or invoice entries, the cross-table records this direct relationship and adds a source pointer. For scenarios without a direct relationship but within the same subject and stage time window, the cross-table records a weak association and marks the location of this weak association as time overlap or subject consistency. Furthermore, the system incorporates field constraints from the unified dimensional / caliber dictionary into the field descriptions of the three-dimensional index to ensure the semantic and unit-level stability of the data fields subsequently collected and transmitted. The system also generates a playback pointer for each three-dimensional index entry, recording the path position for replaying the original record after subsequent collection, used for consistency verification and traceability. During the index binding process, the system sets up a conflict detection strategy to detect situations such as contract number conflicts, duplicate equipment tag numbers, and reused document numbers. When a conflict is detected, a conflict annotation field is added without changing the original value, and the entry is marked in the anomaly registration table. The anomaly registration table will be referenced in subsequent verification and robust denoising processing. Through the above processing, this step outputs the probe primary key index structure, which records the three-dimensional index entries, mapping paths, source pointers, field constraints, and playback pointers.

[0034] In one optional implementation, the index structure generated in step S100 can be generated through a dynamic time-sliced ​​indexing method. That is, after the three-dimensional index binding of contract number, equipment tag number and invoice number is completed, the entire life cycle of the project is dynamically divided into variable-length time slices according to the milestone events of the five stages of investment, construction, grid connection, operation and decommissioning. Each slice independently generates a sub-index with a stage version stamp.

[0035] In another optional implementation, the index structure generated in step S100 can be obtained through a business semantic enhancement indexing method. That is, when binding the main primary key, business semantic tags such as investment and construction mode, transaction category, and grid connection voltage level are extracted simultaneously and encoded as feature bits embedded in the index key.

[0036] In this embodiment of the application, step S200 involves collecting multi-source data on contracts, equipment, and finances based on the probe primary key index structure, and preprocessing the collected data to generate a standard data mirror and anomaly event records, including the following steps B1-B3: B1: Based on the acquisition channels and triggering conditions defined in the probe primary key index structure, acquire contract execution records, equipment operation and maintenance records, and financial invoice records in parallel.

[0037] It should be noted that during the implementation of this step, the probe primary key index structure output above is used as the only entry point. After receiving it, the system first completes session initialization and loading of the acquisition window. The acquisition window is expanded according to the phase time range in the subject-phase mapping table, and an acquisition session queue is constructed for each dimension primary key.

[0038] Specifically, for the contract-transaction probe, the system establishes a read-only session between the transaction platform interface and the contract archive repository according to the contract number and settlement cycle. Records are extracted from the contract terms, performance status, and quotation / settlement fields according to the sampling interval. During extraction, the system adds a source identifier to each contract record based on the source pointer. When supplementary agreements or changes occur, the system loads the corresponding version of the terms image according to the version label in the index and places it into the same session queue, forming a time-ordered contract execution segment. For the equipment-maintenance probe, the system establishes a read-only session between the equipment operation library and the maintenance work order library according to the equipment tag number and stage event number. Records are extracted from channels such as power, energy, status, alarms, maintenance work orders, and inspection records according to the sampling interval. When a status change, alarm trigger, or maintenance work order change trigger condition is identified, the system adds high-frequency extraction and labels the trigger reason and trigger time, forming operation and maintenance segments aligned with the event window. For financial-invoice probes, the system establishes a read-only session between the financial voucher database and the image archive database based on the invoice number and accounting period. Records are extracted from channels such as accounting amount, tax rate, account number, and image pointer according to the sampling interval. When trigger conditions for invoice entry, reversal, or adjustment are identified, the system supplements the relevant voucher entries and binds the image pointer to the invoice number, forming a replayable invoice fragment. Furthermore, the system implements parallel acquisition scheduling for the three types of probe session queues. The scheduling rules follow the priority of the three-dimensional index entries, prioritizing key entries that cross subjects or stages. The replay pointer is updated in real time during acquisition, enabling subsequent consistency verification to quickly locate the source record. If an interface timeout, missing field, or unavailable source is detected during acquisition, the system generates a temporary placeholder entry according to the exception recording strategy and records the exception type, exception time, and affected primary key. The temporary placeholder entry does not participate in value inference but serves as input clues for subsequent verification. After completing the above parallel acquisition, the system aggregates contract fragments, runtime fragments, maintenance fragments, and invoice fragments, reorganizing them into a time-sorted original entry sequence according to the session number and primary key dimension, forming the probe acquisition raw data packet. The probe-collected raw data packet consists of three parts: an entry header, a field body, and a context comment. The entry header carries the primary key, source identifier, and collection window number; the field body carries the extracted field values; and the context comment carries the trigger conditions and version markings. It is important to emphasize that the output field name for this step is "Probe-collected Raw Data Packet," and this raw data packet will be used as input for subsequent steps, entering the source signature verification and clock alignment process through the session number and primary key dimension.

[0039] B2: Perform source signature verification and clock alignment processing on the collected multi-source data.

[0040] It should be noted that during this step, the output probe-collected raw data packets are used as the sole input. The system first retrieves the source summary based on the source identifier and playback pointer in the entry header. For contract execution records, the system reads the record summary from the contract archive repository and transaction platform interface, and matches the summary with the source identifier in the entry header. If the match is consistent, the matching result is recorded and the system proceeds to field consistency verification. For grid-connected metering records, the system reads the metering summary from the grid-connected metering interface and matches it with the collection window number in the entry header. If there are metering entries spanning multiple windows, they are split into multiple segments according to chronological order, and a splitting flag is added to the entry header. For maintenance events, the system reads the event summary from the equipment operation library and maintenance work order library, and matches it with the event window in the entry header. If they are inconsistent, the event summary is placed in a delay queue, waiting for subsequent window shifts before matching. For invoice entries, the system reads the voucher summary and image summary from the financial voucher library and image archive library, and matches them with the invoice number in the entry header. If there are multiple versions with the same number, all versions are loaded according to the version label, and a version sequence field is added to the field body. After the source summary is loaded, the system enters the source signature verification process. The source signature verification merges the entry header, field body, and source summary into a verification fragment according to the field order, and calls the source pointer to access the source verification information. The verification results are divided into three categories: consistent, missing items, and conflict. For consistent entries, the system marks the field body as verified. For missing items, the system adds missing item annotations and missing item sources without modifying the field values. For conflicting entries, the system records the conflict source and conflict field, and marks the entry in the anomaly candidate list. After source signature verification is completed, the system enters the clock alignment process, which maps time stamps from different sources to a unified time anchor. The system extracts the occurrence time, posting time, settlement start and end time, and event time from the entry header and source summary according to the subject-stage mapping table and time anchor table, and constructs a time mapping relationship to unify cross-source time fields onto the stage time axis. If an offset is found between the time field and the stage time axis, the system marks the entry as early, late, or spanning periods according to the offset type and records it in the alignment annotation. For entries with multiple time fields, the system prioritizes retaining the combination of occurrence time and settlement time as the primary alignment reference for the entry, and retains the reference paths of other time fields in the alignment annotation to maintain playback capability. After source signature verification and clock alignment are completed, the system structurally organizes the entries, writing the verified fields into the main field area, writing missing or conflicting fields into the pending processing area, and writing the alignment annotation into the context annotation area. Simultaneously, the system generates a verification fingerprint for each entry, recording the source signature verification result, time alignment status, and field processing status for subsequent robust processing. Finally, the system re-aggregates the structured entries by session number and primary key dimension, and stores them in a segmented buffer container to form a verified data buffer.It should be emphasized that the verified data buffer is explicitly defined as an output field name in this supplement, and is called as an input by the verified data buffer in step B3 to perform missing test labeling and robust denoising of abnormal segments.

[0041] B3: Perform missing test labeling and abnormal segment identification on the checked and aligned data, and generate a standard data image containing the cleaned data based on the identification results, as well as an abnormal event log that records abnormal information and processing suggestions.

[0042] It should be noted that during the implementation of the steps, the verified data buffer output is used as the only input. The system first expands the item sequence in the segmented buffer container according to the session number, primary key dimension and stage time axis. For each item, the system checks the missing item markings in the pending area. The missing item markings are due to the source being unavailable, the field not being filled, or the missing items spanning different periods. The system follows the principle of only marking missing items, without performing numerical estimation or filling. It only retains the identification of the missing item type, missing item source and missing item impact range on the item, and adds the item to the missing item tracking list. The missing item tracking list is used as a reference for weight and allocation rules in the subsequent path calculation, without changing the original field value range. Subsequently, the system enters the abnormal fragment identification and robust denoising process. Abnormal fragment identification is based on a comprehensive judgment of fragments in the entry sequence that do not meet the conventional boundaries, using alignment annotations, verification fingerprints, and context annotations. Specifically, the system identifies abnormal settlement differences in contract execution records, abnormal jumps and abnormal outages in grid connection metering records, abnormal density and long-term status in operation and maintenance events, and abnormal definitions and cross-period accounting in invoice entries. For the identified abnormalities, the system merges adjacent entries and context annotations into an abnormal fragment structure, and records the primary key, session number, stage time, abnormal type, and source reference path on the abnormal fragment. The abnormal fragment structure is then added to the list of pending abnormalities. Robust denoising performs fragment-level noise reduction on abnormal segments without altering the original field values. Specifically, for abnormal jumps and outages, the system uses event window segmentation to split the entries before and after the abnormality, and applies soft boundary indentation to the split segments in time. Soft boundary indentation involves slightly shifting the segment boundaries on the stage time axis so that the cutting point falls near the event time recorded by the triggering condition. For dense anomalies and persistent states, the system inserts interval markers between adjacent entries to remind subsequent path calculations to use segmentation rules in that segment. For abnormal definitions and cross-period accounting, the system moves the entries to a separate definition verification queue and records the definition revision suggestion path on the entries. The suggested path is not executed in this step but is used as the basis for revision when the field is solidified later. After robust denoising, the system extracts the abnormal segment structure from the list of anomalies to be processed, compiles an abnormal event queue, and groups the abnormal event queue by primary key and stage time, recording the anomaly type, triggering condition, soft boundary indentation information, and suggested path. A cross-step reference pointer is attached to the head of the queue so that subsequent field solidification can directly access the original context. At the same time, the system performs a one-time mirroring of the robustly processed sequence of entries. The mirroring process copies the contents of the main field area, the pending processing area, and the context comment area to a read-only container without changing the order of entries or the content of the fields, forming a data buffer mirror. The data buffer mirror is used for subsequent field solidification and primary key mapping to ensure that the status of the entries referenced in subsequent processes is consistent with the status of the entries at the end of this step.Ultimately, the output fields of this step are named Exception Event Queue and Data Buffer Mirror. The Exception Event Queue and Data Buffer Mirror are read by C1 as the input and exception reference of the Data Buffer Mirror, respectively, for field solidification and primary key mapping processing. From the perspective of cross-main steps, the Exception Event Queue also serves as a reference clue for the allocation rules when constructing the subsequent cost transmission path, but does not directly trigger any path revision in this subsection.

[0043] It's important to further explain that the robust denoising method in this step is fundamentally based on constructing a complete data processing framework of identification, classification, labeling, and preservation. This framework enables intelligent identification and robust processing of abnormal data fragments while fully preserving the integrity and traceability of the original data. Specifically: First, the verified data is scanned using a multi-modal anomaly identification algorithm, including: anomaly jump identification based on sliding window Z-score detection (such as contract settlement discrepancies and power surges), and anomaly outage detection based on state duration. The system can automatically identify various anomaly types, such as jumps, outages, dense events, and persistent states, and records key attributes such as type, confidence level, and time range for each anomaly fragment. Second, the system employs differentiated robust processing strategies for different anomaly types. For jump and outage anomalies, a soft boundary indentation algorithm is used to intelligently adjust the boundaries of the anomaly fragments to the most recent event time point, achieving boundary optimization. For dense events and persistent states anomalies, an interval label insertion algorithm is used to insert warning labels in high-density data areas, prompting the subsequent calculation engine to perform segmented processing. All processing adheres to the principle of non-destructiveness, meaning that the original measurements are not modified. Instead, processing recommendations are recorded by adding metadata (such as soft boundaries, interval markers, and adjustment vectors). Ultimately, two key outputs are generated: an anomaly event queue, which fully records all identified anomalies and their processing paths, providing a basis for subsequent analysis and tracing; and a data buffer mirror, a clean copy of the original data and all processing metadata, which can be directly used for downstream computation.

[0044] In one optional implementation, the generation of standard data mirror and abnormal event record in step S200 can be achieved through an incremental difference snapshot mechanism. That is, during probe collection, only the changed data relative to the previous verification period (such as contract clause revision, measurement value jump, new invoice entry) is captured. The difference fragments are quickly identified by hash fingerprint comparison, and the difference content is packaged into a snapshot incremental package with a timestamp. Anomaly detection focuses on the mutation difference rate between adjacent snapshots rather than an absolute threshold. When the difference rate exceeds the dynamic baseline, an abnormal event record is generated, and a snapshot version chain is maintained for subsequent version-by-version backtracking.

[0045] In another optional implementation, the generation of standard data mirror and abnormal event records in step S200 can also be achieved through a multi-source confidence weighted fusion mechanism. That is, for the same source fields (such as settlement electricity) of contract, metering, and financial multi-source probes under the same cost unit, a unified observation model is constructed. The fusion weights are dynamically allocated according to the historical accuracy, collection timeliness, and interface stability of the source system (such as financial voucher weight 0.5, grid-connected metering weight 0.3, and contract record weight 0.2). A weighted average is then performed to generate a standard data mirror with a confidence score. Abnormal event records are no longer independently labeled according to the source, but are generated based on multi-source consistency deviation (such as triggering anomalies when the mutual deviation of the three source data is >15%), directly solving the problem of inconsistent standards at the data entry point.

[0046] In this embodiment of the application, step S300 involves data normalization processing based on standard data mirroring, identifying cost occurrence points, and constructing a dynamic cost transmission path network by combining predefined allocation rules and triggering conditions. Costs are then aggregated along the path network to generate project-level cost summary data, including the following steps C1-C3: C1: Performs dimensional unification and scope boundary processing on the fields in the standard data mirror to complete data normalization, and maps the normalized data to multiple predefined cost units to form a cost unit dictionary.

[0047] It should be noted that during this step, the output data buffer mirror is used as the sole data source, and the output unified dictionary of units / calibers is used as the constraint basis for field definitions and unit calibers. The system first reads the main field area, pending processing area, and context annotation area in the data buffer mirror segment by segment at the data access layer, and loads the corresponding time window according to the phase time range of the subject-phase mapping table. After loading, the system enters the field solidification process. Field solidification refers to the unification of mixed fields from contract execution records, grid connection metering records, operation and maintenance events, and invoice entries according to the field names, field meanings, value ranges, and unit specifications given by the unified dictionary of units / calibers. Specifically, the system performs name merging on fields with the same name but different sources, performs alias mapping on aliased fields, performs unit conversion on fields with inconsistent units, performs value boundary trimming on fields with different calibers, and adds field solidification marks to the field body. For missing fields in the pending processing area, the system only retains the missing item label, source, and scope of impact, without performing numerical estimation or interpolation. For alignment comments, trigger conditions, and version labels recorded in the context comment area, the system collapses them into searchable comment pointers, maintaining their association with the entries for segmentation and version selection during subsequent path construction. After the fields are solidified, the system enters the primary key mapping process. Primary key mapping refers to mapping the three-dimensional primary key, with the contract number, equipment tag number, and invoice number as its core, to a predefined cost unit field. The cost unit domain is categorized into six types: owner-side, electricity-side, grid connection and trading, energy storage and flexibility resources, compliance and carbon certificates, and decommissioning and remediation. Each type of cost unit has an independent set of fields and mapping rules. Specifically, the system maps contract execution records to owner-side and electricity-side cost units based on the entity code, contract number, and stage time. When contract terms involve grid connection fees or trading fees, the mapping extends to the grid connection and trading cost unit. The system maps equipment operation and maintenance entries to energy storage and flexibility resource cost units based on equipment tag numbers and event windows. When maintenance entries include decommissioning or replacement events, the mapping extends to the decommissioning and remediation cost unit. The system maps invoice entries to the corresponding entity's financial accounting aspect based on invoice number and accounting period, and allocates entries to the corresponding cost units according to account mirroring. During the above mapping process, for entries with a caliber verification queue mark, the system only records the mapping relationship and annotation pointer, without revising the caliber. For entries with weak correlations, the system records both the weak correlation mark and the basis for its formation (overlapping time or consistent entity). After the primary key mapping is completed, the system integrates the solidified fields with the allocated cost units to generate structured cost unit entries. The cost unit code, main code, stage time, and source overview are written in the entry header, and the solidified field values ​​and comment pointers are written in the entry body.Finally, the system archives items under the same cost unit according to stage time, forming a cost unit dictionary. The cost unit dictionary contains three parts: cost unit code index, field definition index, and item archiving index, which can be used for unit-by-unit traversal when assembling allocation rules and triggering conditions. Finally, the output field of this step is named Cost Unit Dictionary. The cost unit dictionary is called as input by step C2 and enters the allocation rule and triggering condition assembly process through cost unit code and stage time. At the same time, this output is indirectly referenced as the pre-data field of the S400 scenario assembly and calculation engine and the comparison output when crossing main steps.

[0048] In an optional implementation, when multiple triggering conditions act simultaneously in step C1, the system determines the triggering order according to the priority table through a forced sorting method based on precise rule matching. That is, the system directly extracts the specified triggering order and generates an unchangeable deterministic rule, ensuring that mandatory constraints such as default clauses taking precedence over normal settlement and asset impairment taking precedence over regular depreciation are strictly enforced. Example 1: A contract might stipulate: "When an unplanned outage occurs (alarm triggered), the current period's electricity settlement will be recalculated according to the default clause and will take precedence over the normal settlement clause." The system will generate a rule accordingly: Alarm Trigger (Unplanned Outage) -> Triggering Condition Priority = High. Example 2: Accounting standards require that "the recognition of asset impairment (a change in accounting period) should take precedence over regular depreciation in accounting treatment." The system will assign a higher priority to asset impairment events.

[0049] In another optional implementation, when multiple triggering conditions act simultaneously in step C1, the system can determine the triggering order based on a priority table using a flexible sorting method with a default priority table. This allows users to dynamically adjust the relative weights of common triggering conditions according to management strategies such as cost priority or risk aversion priority. When no mandatory rule matches, the system flexibly executes the triggering conditions according to this configurable default order, recording the delayed conditions as alternatives, thus achieving a balance between decision-making flexibility and traceability. Example 3: If risk aversion priority is selected, the priority of triggering conditions related to risk sharing and insurance payouts will be automatically increased.

[0050] C2: Based on the cost unit dictionary, identify the time points corresponding to cash flow, physical flow, or voucher flow as cost occurrence points; load the corresponding allocation rule templates for the identified cost occurrence points and bind the rule parameters to the cost unit fields; based on the bound allocation rules and associated triggering conditions, construct a directional cost transmission path network.

[0051] It should be noted that during the implementation of this step, the cost unit dictionary is used as the sole input. The system first opens the corresponding archive based on the cost unit code index and traverses the entries to identify cost occurrence points. A cost occurrence point refers to the cost record position on the timeline corresponding to cash flow, physical flow, or document flow. Its definition is jointly determined by the fixed field markers in the entry body, the event type, the accounting period, and the contract settlement cycle. Specifically, the system marks settlement entries, default entries, and price adjustment entries in the contract execution record as contract-related cost occurrence points; marks metering entries related to capacity fees, transmission and distribution fees, or ancillary services in the grid connection metering record as grid connection and transaction-related cost occurrence points; marks replacement, repair, and decommissioning / reactivation entries in the equipment operation and maintenance entries as energy storage and flexibility resource-related cost occurrence points; marks accounting, reversal, and adjustment entries in the invoice entries as financial-related cost occurrence points; and marks entries related to decommissioning events as decommissioning governance-related cost occurrence points. For entries marked with a caliber verification queue or weak correlation marker, the system provides an event annotation pointer for segmentation or weight reduction during subsequent path calculations. After identifying the cost incurrence point, the system enters the allocation rule assembly stage. Allocation rules are structured rules used to allocate the amount, quantity, or time period of a single cost incurrence point to one or more responsible entities, cost objects, or settlement channels. The system loads rule templates according to cost unit type and binds the parameter bits in the templates to the cost unit fields. For example, the allocation rule templates for the owner side and the electricity user side include four categories: fixed allocation, proportional allocation, tiered allocation, and risk sharing. When binding, the basic unit price, profit-sharing ratio, tiered threshold, and risk-sharing agreement from the contract terms are written into the rule body. The allocation rule templates for grid connection and transaction stages include three categories: capacity-related, electricity-related, and event-related. When binding, the grid-connected capacity, billed electricity, and event window are written into the rule body. The allocation rule templates for energy storage and flexibility resources include three categories: time-based allocation, performance coefficient allocation, and event-triggered allocation. When binding, the operating period, performance evaluation score, and triggering event are written into the rule body. The allocation rule templates for decommissioning and remediation include two categories: asset residual value mapping and disposal cost allocation. When binding, the equipment replacement list and disposal plan summary are written into the rule body. Subsequently, the system enters the trigger condition assembly stage. Trigger conditions refer to external or internal conditions that cause a rule to take effect, change, or terminate. The system loads candidate trigger conditions from the context comment pointer and aligns them with contract status changes, alarm triggers, maintenance work order changes, and accounting period changes. When multiple trigger conditions are active simultaneously, the system determines the trigger order according to a priority table (not a static configuration, but a set of decision logic dynamically generated and managed based on business rules) and records lower-priority trigger conditions as alternative triggers for selection during path calculation.After the allocation rules and triggering conditions are assembled, the system generates directed edges based on the cost occurrence point, allocation rules, and triggering conditions. Using the main body code, cost unit code, and stage time as node keys, and the allocation rule body and triggering condition body as edge attributes, the system constructs paths from generation → collection → transmission → allocation, forming a cost transmission topology (cost transmission path network). During topology construction, the system sets segment nodes for segments within the coverage area of ​​the abnormal event queue. These segment nodes mark soft boundary indentation information and suggested paths in their edge attributes. Edges on the paths do not modify fields; they only carry comments and control parameters. Finally, the system serializes the topology into a cost transmission-allocation path graph, which consists of three parts: a node table, an edge table, and a rule table. The node table records instances of the subject and cost unit in the stage time, the edge table records the allocation relationship and triggering conditions, and the rule table records the binding details and version labels of each template instance. The output field of this step is named Cost Transmission-Amortization Path Graph. The Cost Transmission-Amortization Path Graph is called as input in step C3 and enters the collection and path calculation process based on the node table and edge table.

[0052] C3: Perform cost value aggregation calculations along the cost transmission path network to generate project-level cost summary data.

[0053] It should be noted that during the implementation of this step, the output cost transmission-allocation path graph is used as the sole input. The system first establishes a calculation session based on the node table and edge table, traverses the node instances within each stage's time window, and performs aggregation calculations according to the allocation order of the edge table. Aggregation processing refers to accumulating the amount or quantity at the cost occurrence point to the target node instance according to the allocation rules and triggering conditions, without rewriting the original entries, along the direction of generation → aggregation → transmission → allocation. During execution, the system reads the rule body and version label in the rule table. When encountering an edge with alternative triggers, it determines whether the triggering conditions are met one by one according to the triggering order, and records the effective trigger and triggering time when the conditions are met. When encountering segmented nodes, the system divides the path into multiple calculation segments according to the soft boundary indentation information, and performs aggregation separately for each calculation segment. The aggregation results are merged at the end of the segment. When encountering weakly related edges, the system adds weak association markers and formation basis without changing the amount or quantity, and the aggregation results enter the subsequent interpretation domain. After the aggregation process is completed, the system performs path calculation for each target node instance. Path calculation refers to merging, deduplicating, and selecting versions of multiple source paths aggregated to the same node instance, and outputting node-level cost entries. During the merging process, the system arranges path segments according to time order and version label. For contract-type paths with version replacement or supplementary agreements, version selection is performed. Version selection uses time coverage relationship (the core rule is that the later effective version covers the earlier effective version, and the later effective version is given priority in the overlapping interval) to determine the priority segment, and the covered segment is written into the version replacement annotation. During the deduplication process, the system identifies duplicate entries from the same source based on the source identifier and field fixed mark, retains the first entry as the master record, and writes subsequent duplicate entries into the duplicate annotation. After the node-level entries are generated, the system groups the node-level entries according to the subject code and cost unit code to form subject-level and unit-level aggregation tables. To facilitate subsequent scenario calculations and cash flow mapping, the system generates a phase summary on the aggregation table. The phase summary outputs fixed fields such as amount, quantity, billed electricity, capacity-related fields, and event counts according to the phase time, and includes a summary of trigger conditions and anomaly reference pointers. The system also outputs an explanatory domain, which records version replacement comments, duplicate comments, weak association markers, and segment node information. The explanatory domain does not participate in numerical calculations and is only used for subsequent visualization and strategy suggestion generation. After the path calculations for all node instances are completed, the system enters project-level assembly. Project-level assembly refers to unifying the aggregation tables at the subject and unit levels into a project-side summary structure, organizing entries in the summary structure according to the subject-unit-phase hierarchy. The system merges the phase summaries at each level into a project-level phase summary, merges the explanatory domains into a project-level explanatory domain, and writes the project name, project code, and phase time range at the header of the summary structure.Finally, the system outputs a project-level cost aggregation package, which consists of three parts: a summary structure, a project-level phase summary, and a project-level interpretation domain. The package header records a cross-step reference pointer, pointing to the field entry required in step D1. The project-level cost aggregation package is directly input and called as a subsequent project-level cost aggregation package for scenario matrix assembly and cash flow mapping. At the same time, this output is referenced in the visualization and strategy suggestion stages across main steps to generate graph interpretations and allocation explanations.

[0054] Preferably, this step completes the hierarchical aggregation and path calculation from nodes to projects under the constraints of rules and triggering conditions, forming a structured, interpretable and replayable project-level cost aggregation package, and outputs the explanatory information and stage summary together, which is convenient for direct use in subsequent scenario accounting and visualization.

[0055] Specifically, in step S300, field values ​​are normalized using formula ① to unify multi-source field values ​​to within the standard units and scope boundaries. Formula ①: For each standard field f, calculate the unified value. for: Formula① In the formula: It comes from the data buffer mirror. fields The original value; It is a source in the dictionary of unified dimensions / calibers. fields Dimension conversion factor; and It is a field in the unified dictionary of dimensions / calibers The minimum and maximum caliber boundaries; It has fields The source set; It is the cardinality of the set; It is a field The unified value; : Field identifier; Source index; Minimum value function; : The function to find the maximum value.

[0056] It should be noted that the acquisition of the unit conversion factor is automatically completed by the system by parsing the data source metadata and calling the built-in unit conversion library. Manual intervention and confirmation are only required for complex conversions or when the system cannot automatically determine the conversion. The minimum and maximum caliber boundaries are not defined by a single role, but rather reflect various business rules such as contracts, technology, policies, and management requirements. That is, by parsing relevant documents and accepting user input, these rules are quantified into specific numerical boundaries and entered into the unified unit / caliber dictionary.

[0057] It is understandable that data source mapping involves extracting the original field values ​​from the primary field area mirrored by the data buffer. Conversion factors are extracted from a unified dictionary of dimensions / calibers. With boundary , Together, they form the input in formula ①; formula ① As the standard value after the field is fixed, it is used for subsequent cost unit entry construction.

[0058] Furthermore, by mapping the primary key to the cost cell domain using formula ②, the three-dimensional primary key is associated with six types of cost cell codes, namely, cost cell codes. From the primary key feature vector Determined by linear mapping and normalization function: Formula② In the formula, σ(x) is the sigmoid function; It is a numerical mapping of the main encoding; It is the hash value of the contract number; It is the numerical code of the device tag number; It is the integer representation of the ticket number; is the normalized timestamp value of the stage time; T: transpose symbol; w is the weight vector; u is the cost unit code, an integer from 1 to 6, corresponding to the six types of cost units respectively.

[0059] It is important to understand that the main body encoding The core purpose of numerical mapping is to transform discrete, non-numerical entity identifiers (such as owner company A and electricity-consuming company B) into continuous numerical features with business meaning, so that the subsequent mathematical model (formula ②) can handle and learn the impact of different entities on cost unit attribution. Furthermore, the weight vector... The replication is not a simple, indiscriminate copy. It is a guided, traceable initialization process: its core is templating: ensuring consistency in the classification logic benchmark for similar cost units across different projects; allowing project-level calibration: enabling the model to adapt to the specific business characteristics of each project; and following phase inheritance: ensuring the stability of cost aggregation logic throughout the project lifecycle. This approach, while guaranteeing algorithm consistency and repeatability, also provides the necessary flexibility, making it a key design feature for achieving precise calculations and traceability.

[0060] Data source mapping: This consists of primary key information extracted from the entry headers mirrored in the data buffer. Weights are extracted from a unified dictionary of dimensions / calibers. Formula ② As a cost unit code, it is used to identify the cost unit to which an item belongs. Formula ① yields... The result obtained from formula ② Together, they are used to construct cost cell entries and output a cost cell dictionary.

[0061] Furthermore, cost incurrence points are identified using Formula ③, transforming items into amortizable cost values. Formula ③: For a cost unit item, the cost incurrence point value is calculated as follows: Formula③ In the formula, It is an indicator function; e i E is the event type code for entry i; E is the set of cost occurrence point event types, including settlement, default, price adjustment, etc., defined by the field definition index of the cost unit dictionary; This is an entry. The amount or quantity value is determined by formula ①. Extract from; It is the cost incurrence point value.

[0062] Data source mapping: extracted from the entry body of the cost unit dictionary Extracted from context comment pointer Formula ③ The base value used as the cost incurrence point is used for subsequent allocation calculations. Formula ④ is used for assembling allocation rules and defining the edge weights between nodes. Formula ④: For node pairs edge weight Determined by the sharing rule template and triggering conditions: Formula④ In the formula, r ij This represents the basic allocation ratio; i and j are node indices, corresponding to cost unit instances; r ij It is the allocation ratio derived from the rule table; It is an indicator function that indicates the fulfillment of the trigger condition, taking the value 0 or 1, determined by the priority table in the trigger condition body; A ij It is the edge weight.

[0063] Data source mapping: extracting r from the rule table ij The trigger state is extracted from the trigger condition body; A in formula ④ ij As an edge attribute, it is used to construct the topology. Equation ③ yields... Used for weight calculation in formula ④, A in formula ④ ij Used for subsequent aggregation.

[0064] Furthermore, the cost is accumulated along the path using formula ⑤. Formula ⑤: For target node j, the accumulated cost C jThe calculation is as follows: Formula⑤ In the formula: It is the set of source nodes pointing to node j; i is the index of the source node; It is the cost occurrence point value in formula ③; A ij The edge weights are those in formula ④; C j It is the cumulative cost of node j.

[0065] Data source mapping: Extracting A from the side table of the cost transmission-allocation path graph ij Extract node relationships from the node table; C in formula ⑤ j The aggregated result is used for path calculation. Formula 6 is used for version selection in path calculation, aggregating multiple version values ​​through attention weights. Formula 6: For node j, the final value... The calculation is as follows: Formula⑥ In the formula: v is the version index; v j It is a set of node versions; It is the cumulative cost across versions; It is attention weight; It is the node-level cost value.

[0066] It should be noted that the attention weight generation logic is a configurable two-dimensional dynamic weighting mechanism: First, using time coverage as an objective benchmark, the proportion of exclusive days for each version within the node time window is accurately calculated and overlapping segments are handled to ensure that the weight allocation conforms to the actual effective duration; second, business importance is introduced as a subjective adjustment, assigning differentiated importance scores based on version type (main contract, supplementary agreement, etc.) to highlight the priority of key business agreements; finally, the two are normalized and multiplied through a combination strategy to form a fine-grained weight vector that respects both historical facts and business intentions, and supports flexible switching of strategies according to cost unit type. This ensures the mathematical rigor of weighted aggregation and realizes the evolution of cost calculation from mechanical arithmetic to intelligent calculation that carries business logic.

[0067] Data source mapping: Extracting version information and time overlap from the project-level interpretation domain; z in formula ⑥ j As the output of path calculation, it is used to generate a project-level cost aggregation package. The C obtained from formula ⑤ j Version aggregation for formula ⑥, z of formula ⑥ j Used for building phase summary.

[0068] That is, by solidifying fields and mapping primary keys, multi-source data is unified into a cost unit dictionary, providing structured input for the assembly of allocation rules; by identifying cost occurrence points and constructing graphs, a traceable allocation path is formed, supporting aggregation calculation and version selection; finally, a project-level cost aggregation package is output to ensure the data consistency and interpretability of cost accounting throughout the entire life cycle.

[0069] In an optional implementation, the construction of the dynamic cost transmission path network in step S300 can also be achieved through a hierarchical transmission model based on a state machine. That is, by defining the cost-bearing status of each responsible entity in different time segments as state nodes, the allocation rules and triggering conditions are transformed into state transition functions. When the cost occurrence point is identified, the system drives the cost value in the source state node according to the bound rules and transfers it to multiple target state nodes according to a predetermined logic. The hierarchical transmission mechanism realizes the aggregation of costs within and between units. During the entire transmission process, each state transition generates a detailed log. The collection of all transmission logs dynamically constitutes a complete cost transmission path network, clearly recording the complete flow trajectory of costs.

[0070] In another optional implementation, the construction of the dynamic cost transmission path network in step S300 can also be achieved through an event-driven cost flow model. That is, each cost occurrence point is encapsulated as a standardized cost event object, and multiple pre-registered rule processors form a processing topology based on their subscription relationships. After the cost event is published, it will flow in the processing topology according to the rules. Each processor processes the event (such as splitting or forwarding) and generates path edges, while the triggering condition serves as a routing policy to determine the event flow direction. Finally, the entire processing topology dynamically constructs a complete cost transmission path network in real time from the processing logs of the cost event flow, realizing the traceability and dynamic response of the cost transmission process.

[0071] In this embodiment of the application, step S400 involves binding project-level cost summary data with external scenario parameters to generate multiple calculation scenarios. For each scenario, a full lifecycle economic accounting is performed to obtain a set of economic indicators. Multi-objective optimization analysis is then conducted on the economic indicator set, and decision support information is synthesized to generate a structured decision package. This includes the following steps D1-D3: D1: Bind the assemblable domain in the project-level cost summary data with the parameter bits in the external scenario parameter set, and generate multiple calculation scenarios by combining the parameter bits.

[0072] It should be noted that during the implementation of this step, the output project-level cost aggregation package is used as the input to the calculation domain, and the scenario parameter set output from step A2 is used as the input to the configuration domain. The system first loads the summary structure, project-level phase summary, and project-level interpretation domain at the assembly entry point, and enables a unified time axis based on the subject-phase mapping table. Specifically, the system performs itemized parsing on the core parameters in the scenario parameter set: direct power purchase price mechanism, power purchase and sale contract structure, energy storage capacity configuration scheme, and tax parameters, as well as optional parameters: green certificate price range, carbon price range, and subsidy parameters. Itemized parsing refers to decomposing text or enumerated configurations into traversable parameter bits and value sets, and writing parameter bit identifiers, value lists, and priority sequences into the parsing results. Furthermore, the system performs an assemblable domain scan on the summary structure in the project-level cost aggregation package. An assemblable domain is a set of fields in the project-level phase summary that can be referenced by scenario parameter bits, covering at least amount-related fields, quantity-related fields, billed electricity-related fields, capacity-related fields, and event count-related fields. When a field with a direct mapping relationship to a scenario parameter bit is scanned, the system records the mapping path and establishes a binding relationship between the parameter bit and the field. When a field requiring an interpretable domain to participate in the selection is scanned (e.g., a version substitution comment or weak association marker exists), the system adds an interpretable pointer to the binding relationship. (An interpretable pointer is a lightweight, structured metadata object (or instruction package) that does not directly store data values ​​but stores a method or path for reliably, consistently, and traceably obtaining a final data value in a complex environment with multiple versions and abnormal data; it is a key technical component for achieving the two major characteristics of interpretability and consistency in this invention.) This is used to maintain a consistent value retrieval path under the same scenario. Subsequently, the system proceeds to the discount rate / WACC configuration binding process. The discount rate refers to the set of discount rules used for time value in the full lifecycle cash flow mapping, including the discount base date, discount step size, and discount method. WACC is the abbreviation for Weighted Average Cost of Capital. WACC configuration binding means storing the cost of capital parameters and the discount rate together in the scenario configuration body, so that subsequent cash flow mappings can directly use a unified cost of capital rate and discount rules when invoked. To this end, the system extracts the relevant parameter bits for cost of capital, tax and subsidy parameter bits from the scenario parameter set and performs integrated binding with the discount rate field. When there are multiple sources of funds or differentiated cost of capital, the system generates multiple cost of capital segments for different regions within the same scenario and writes the applicable time range and subject range for each region in the configuration body.After binding is complete, the system launches the scenario matrix assembler. The assembler generates scenario entries using Cartesian combinations of parameter bits, writing the aforementioned binding relationships and explanation pointers into each entry. For parameter bits that cannot be combined (e.g., the mutual exclusion between contract structure and direct power purchase pricing mechanisms), the assembler skips the combination based on priority and records the reason for skipping next to the scenario entry table. Upon completion of assembly, the system performs a consistency check on the scenario entry table. This check is divided into parameter domain consistency and mapping domain consistency. The former checks for mutually exclusive or missing values ​​in parameter bits within the same scenario, while the latter checks for empty paths or missing explanation pointers in the project-level phase summary fields bound to parameter bits. Scenario entries that pass the check are written into the scenario matrix; those that fail are marked as unusable and retained in the table for subsequent debugging. Finally, the output field of this step is named "Scenario Matrix," which is used as input in step D2 for performing full lifecycle cost accounting and cash flow mapping. Simultaneously, the scenario matrix is ​​referenced as an upstream data domain for chart generation and explanatory text in the visualization and strategy recommendation phase across main steps, but is not elaborated in this step.

[0073] D2: For each calculation scenario, perform cash flow mapping on the project's entire life cycle timeline and calculate the net present value (NPV) based on the bound capital cost parameters.

[0074] It should be noted that during the implementation of this step, the output scenario matrix is ​​used as the sole configuration input. The system first opens an accounting session for each scenario item one by one. The accounting session writes the parameter values, discount rate, and WACC configuration binding body into the session header, and loads the summary structure of the project-level cost aggregation package and the project-level stage summary into the session body. Specifically, the system performs constraint initialization for each scenario item in the accounting engine. The constraint initialization includes four types: contract boundary constraints, grid connection capacity constraints, capital cost constraints, and policy boundary constraints. Contract boundary constraints read version substitution comments from the interpretation pointer to determine the effective range of the contract segment, and only perform value taking on the effective range during accounting. Grid connection capacity constraints read the capacity upper limit from the capacity-related fields of the summary structure, and perform upper limit pruning on the capacity-related billing fields during cash flow mapping. Capital cost constraints directly reference the discount rate and WACC configuration binding body. Policy boundary constraints read tax and subsidy parameters from the scenario parameter bits, and perform rule-based processing of tax rates, subsidies, and surcharges on the relevant items during mapping. Subsequently, the system executes a full lifecycle cost accounting process. This process traverses the timeline from the investment phase, construction phase, grid connection phase, operation and maintenance phase, to the decommissioning and remediation phase. For each phase, it calls the corresponding field domains: for the investment and construction phase, it calls amount-related fields and event count-related fields; for the grid connection phase, it calls capacity-related fields and billable electricity-related fields; for the operation and maintenance phase, it calls amount-related fields, quantity-related fields, and event count-related fields; and for the decommissioning and remediation phase, it calls disposal and residual value-related fields. During these calls, compliance checks for constraints and interpretations are performed. To maintain consistency with the project-level interpretation domain, the system introduces an interpretation tracker at each phase. When the interpretation tracker encounters weakly related markers or segment nodes, it writes the value path and comments of that segment's entries into the session-level tracker, which is then archived during phase summaries. After completing the phase traversal, the system performs cash flow mapping. Cash flow mapping aligns the aforementioned phase entries on the time axis to the step size given by the discount rate, and calls the capital cost parameters and tax / subsidy parameters in the WACC configuration binding body for each time slice to form a time-ordered cash flow sequence. When there are multiple capital cost segments in different regions, cash flow mapping switches the capital cost parameters on the time axis according to the applicable phase time range for each region. When there are policy parameter jumps in time, cash flow mapping divides the time slices according to the jump interval of the parameter bits, and performs rule-based processing on the divided time slices respectively. After the cash flow sequence is generated, the system calculates economic indicator entries within the accounting session. Economic indicator entries are a set of indicators obtained by rule-based calculation of the combined fields of the cash flow sequence and phase summary, which at least includes indicators related to the cost per kilowatt-hour and indicators related to net cash flow, and references session-level trajectories in the entries to maintain interpretability.To facilitate subsequent comparisons, the system performs consistency checks on the economic indicator entries at the end of the session. The checks include consistency and readability between the parameter and data fields. If the check passes, the economic indicator entry is written to the session exit, becoming a member of the economic indicator set. If the check fails, the system marks the entry as unreadable and retains its trace for later investigation, excluding it from the comparison set. After each session completes according to scenario entries, the system merges all economic indicator entries from all session exits into an economic indicator set. The economic indicator set records the scenario entry index and explanation index at the set header. Finally, the output field of this step is named "Economic Indicator Set," which is used as input in step D3 for robust sorting, Pareto filtering, and the synthesis of visualization elements. Simultaneously, the economic indicator set is referenced in the standardized report stage across main steps to generate scenario comparison tables and explanatory text.

[0075] D3: Robustly sort the economic indicators set for all scenarios to reduce the impact of data quality defects on sorting stability; perform Pareto front screening on the sorted economic indicators set to identify the non-dominated solution set as the optimal scenario candidate; based on the optimal scenario candidate and its economic indicators, synthesize a structured decision package containing visual comparison elements and explanatory annotations.

[0076] It should be noted that during the implementation of this step, the output set of economic indicators is used as the sole input. The system first loads the scenario entry index and explanation index into the comparison manager, and establishes a dual-view structure of indicator view and explanation view for each economic indicator entry. The indicator view carries the indicator fields used for sorting and filtering, while the explanation view carries the session-level trajectory, weak association markers, version replacement annotations, and segment node information used for explanation. Specifically, the system first performs robust sorting processing, which refers to the process of performing robust processing on possible abnormal situations within the indicator view before sorting. Abnormal situations include missing test annotations, jumps, and scale differences across scenarios for indicator fields. During the robust processing phase, the system performs weight reduction on indicator fields with missing test annotations by reducing the participation weight of the field in the sorting weight set. For indicator fields with jumps, intra-segment normalization and inter-segment concatenation are performed according to the jump trajectory in the explanation view before sorting, so that the sorting criteria are compared under the same semantics. For indicator fields with scale differences, the system uses unit mapping derived from the unified dictionary of dimensions / calibers for unified expression. After robustness processing, the system sorts the economic indicator items in descending or ascending order according to the sorting key, and writes robust sorting annotations into the sorting results, recording the basis for weight reduction and normalization to maintain replayability. After robust sorting, the system performs Pareto screening. Pareto screening refers to identifying a candidate set that cannot be simultaneously dominated by other scenario items in all dimensions across multiple indicator dimensions. The system traverses the items in the indicator view and checks for explicit dominance relationships, marking items belonging to the non-dominant set as Pareto front members, and recording their relative positional relationship with other items in the interpretation view. When multiple items are parallel on certain indicators, the system calls the weak association markers and segmentation node information in the interpretation view as parallel dimensions, providing parallel annotations without hard breaking. After Pareto screening, the system proceeds to the visualization element synthesis process. These visualization elements are embeddable graphical or tabular objects used for standardized reports, including at least a scenario matrix overview, a cost transmission diagram, an economic indicator comparison chart, and a sensitivity ranking illustration. The system extracts parameter bits from the scenario entry index to construct the scenario matrix overview, synthesizes a cost transmission diagram summary from the node and edge tables of the cost transmission-allocation path graph, synthesizes an economic indicator comparison chart from the ranking results and the Pareto front set, and extracts weak correlation markers and segmented node information from the explanatory index as annotation sources for the explanatory text. These graphical or tabular objects are bound to the explanatory view during synthesis, and the bound objects can be clicked to replay their corresponding session-level trajectory during the report generation phase.Finally, the system packages the robust sorting results, Pareto front set, and visualization elements together to form a structured decision package. The structured decision package records the reference relationship between the unit / caliber unified dictionary in the package header and the output, which is used to maintain consistency in caliber during subsequent write-back and iteration. The structured decision package has clearly defined output field names as the main step, and it is input back to the unit / caliber unified dictionary in the closed loop to trigger incremental registration of the dictionary and revision of parameter bits. At the same time, it is directly referenced in the standardized report and strategy suggestion stage across main steps to generate scenario matrix panels, economic comparison tables, and explanatory text.

[0077] Furthermore, this method proposes a refined weight decay mechanism, which replaces the traditional binary processing with a multi-factor continuous weight reduction function. The weights are smoothly adjusted based on three dimensions: the proportion of effective data ρ, the penalty coefficient for missing data type α, and the impact factor of missing data mode γ. ρ directly reflects the proportion of missing data, α is assigned weights differently according to the reasons for missing data such as sensor failure, manual omission, and failure to generate data across periods, and γ quantifies the destructiveness of continuous missing data. The coupling of the three factors enables the weight decay to be precisely matched with the degree of information loss, achieving truly robust weight reduction.

[0078] For data jump scenarios, intra-segment normalization employs an improved quantile scaling method, using the 5% and 95% quantiles within the data segment as normalization boundaries. This replaces the traditional Min-Max method, which is susceptible to extreme outliers, effectively resisting the influence of outliers within segments. This ensures that data in each segment before and after the jump accurately reflects the relative trend at independent scales, laying a robust foundation for subsequent comparisons. Simultaneously, to guarantee the continuity of different normalized segments, an "anchor alignment" mechanism is introduced for inter-segment splicing. Boundaries are anchored using event timestamps such as jump trigger points or stable observations, quantifying inter-segment offsets and shifting the subsequent segment as a whole to ensure numerical continuity at anchor points. All anchor points, offsets, and splicing operations are recorded in the interpretation view, ensuring that the spliced ​​global sequence possesses both trend consistency and process interpretability, ultimately providing high-quality input for robust sorting.

[0079] Preferably, this step performs robust ranking and multi-indicator non-dominant screening on the set of economic indicators, generates visualization elements bound to the explanatory view, forms a structured decision package, and writes back the dictionary reference relationship to support the consistency of subsequent reports and parameter iterations.

[0080] Specifically, in step S400, the mutual exclusion validity indicator function of the scenario item is calculated using formula ⑦ to ensure that the parameter bit combination meets the constraints: For scenario item s, its mutual exclusion validity indicator function I(s) is defined as: Formula⑦ In the formula: s represents the scenario entry; i, j: parameter bit indices; M is the set of mutually exclusive pairs; It is a mutual exclusion check function; Let i be the value of the parameter. The value of parameter j is given; max: the function to find the maximum value; I(s) is the mutual exclusion validity indicator function, with a range of {0,1}.

[0081] Data source mapping: Extracting parameter bit values ​​from the scenario parameter set and mutual exclusion rule M and Formula ⑦ is used as an intermediate quantity for subsequent consistency verification.

[0082] The consistency check score for scenario items is calculated using Formula ⑧, combining mutual exclusion validity and mapping domain integrity: For scenario item s, the consistency check score C(s) is defined as: Formula⑧ In the formula: I(s) is the mutual exclusion validity indicator function of formula ⑦; It is a path integrity indicator function. It takes a value of 1 when there is a non-empty path in the project-level stage summary field that all parameters are bound to, and 0 otherwise. It is derived from the binding relationship of the assemblable domain scan. C(s) is the interpretation pointer validity indicator function, which takes the value 1 when all interpretation pointers exist and are valid, and 0 otherwise. It is derived from the project-level interpretation domain. C(s) is the consistency check score, with a value range of [0,1].

[0083] Data source mapping: Extracted from I(s) in formula ⑦ and path information of the assembleable domain scan. Project-level interpretation domain extraction C(s) in Formula ⑧ serves as the verification basis; when C(s) = 1, the scenario item passes the verification and is written into the scenario matrix. Scenario items that pass the verification are written into the scenario matrix, while scenario items that fail the verification are marked as unusable and kept in the table for subsequent debugging.

[0084] Furthermore, the cash flow value for each time slice is calculated using formula ⑨, and allocated proportionally to the stage cost: For time slice k, its cash flow CF^{k} is defined as: Formula⑨ In the formula: k represents the time slice index; i represents the stage index; S is the total number of stages; This is the cost value for stage i, derived from the amount-related fields in the project-level stage summary; It is the overlap time length between stage i and time slice k; It is the duration of stage i; It is the cash flow value of time slice k.

[0085] Data source mapping: extracted from project-level phase summaries Extracted from the subject-stage mapping table The time period is divided by the discount rate, and the time slices are extracted according to formula ⑨. This serves as an intermediate quantity for subsequent indicator calculations. The Net Present Value (NPV) indicator is calculated using formula ⑩, and is used as one of the economic indicators: For each scenario item, the NPV is defined as: Formula⑩ In the formula: k is the time slice index; K is the total number of time slices; It is the cash flow value of formula ⑨; It is the discount factor for time slice k; NPV is the net present value.

[0086] Data source mapping: by formula ⑨ and discount caliber Calculate NPV using formula 10 as part of the economic indicators.

[0087] Furthermore, the weight adjustment for robust ranking is calculated using formula (11), and the missing test labels are weighted less: For index j, its robust weight w^{j} is defined as: Formula (11) In the formula: j represents the index; β is the base weight of indicator j; β is the weighting factor. It represents the degree of missing measurement of indicator j; It is the robust weight of index j.

[0088] It should be noted that the degree of missing data is achieved through continuous proportional quantification to achieve fine-grained robustness. Its value is determined by the proportion of missing data to the expected total data volume, and it takes a smooth value in the range of [0,1]. For example, missing 3 days in a 30-day cycle is quantified as 0.1, missing 15 days as 0.5, and completely missing as 1. This design makes the weight decay and the severity of missing data exponentially smooth, which avoids the coarseness of traditional binary judgment and can dynamically weaken the ranking influence of unreliable indicators according to the actual data completeness.

[0089] Data source mapping: extracted from the indicator view of the economic indicator set. Extracted from the set of sorting weights Formula (11) It is used as a sorting weight for subsequent sorting.

[0090] Formula (12) is used to calculate the Pareto dominance indicator function for scenario items, which is then used to select frontier members: For scenario items a and b, the indicator function D(a,b) that indicates that item a dominates item b is defined as: Formula (12) In the formula: a and b represent two scenario entries; j represents the index; J is the total number of indicators; and These are the values ​​of index j for entries a and b, respectively; It is an indicator function that takes the value 1 when the condition is true and 0 otherwise; It is the dominance indicator function, which takes the value 1 when a dominates b, and 0 otherwise; ∏: multiplication symbol.

[0091] Data source mapping: extracted from the indicator view of the economic indicator set. and Formula (12) Used to determine dominance; Pareto front members are those that satisfy the condition for all b. Entry a = 0.

[0092] In summary, the beneficial effects of this invention's method for precise calculation of the full life cycle cost of green electricity direct connection projects are as follows: It constructs a closed-loop system for full life cycle cost calculation through the synergistic effect of three major links: First, at the data acquisition end, relying on a three-dimensional index and robust denoising mechanism, it only marks missing data without modifying the original data, thus accumulating abnormal behavior into a traceable event queue, providing stable and reliable data input for subsequent processing. Second, at the cost aggregation end, through a dynamic cost transmission graph, it solidifies the relationships between fields, allocation rules, and triggering conditions into a topological structure of nodes and edges, achieving transparency and version replayability of cost allocation paths for multiple entities and multiple stages, ensuring accurate aggregation under scenarios with differences in caliber and weak correlations. Finally, at the scenario comparison end, with the help of dynamic parameter binding and robust sorting-Pareto screening mechanism, it agilely responds to changes in policies, contracts, and capacity schemes, generating a visual decision package bound to the interpretation view and writing back to the dictionary for iterative optimization. The three elements are interconnected, forming a complete closed loop from data cleaning and cost transmission to intelligent decision-making. This significantly improves the multi-entity collaboration efficiency, cost accounting accuracy, and standardized decision-making capabilities of green electricity direct connection projects in complex market environments.

[0093] Example 3 illustrates a schematic scheme for a precise calculation method of the full life cycle cost of a green power direct connection project. It should be noted that the technical solution of this precise calculation system for the full life cycle cost of a green power direct connection project belongs to the same concept as the technical solution of the aforementioned precise calculation method for the full life cycle cost of a green power direct connection project. Details not described in detail in this example can be found in the description of the technical solution of the aforementioned precise calculation method for the full life cycle cost of a green power direct connection project.

[0094] This embodiment also provides a system for detailed calculation of the full life cycle cost of green electricity direct connection projects, including: The generation module is used to standardize the main body identification and timeline based on the project's basic information and investment and construction mode, as well as to deploy and bind the multi-source data acquisition probes to generate the probe primary key index structure. The preprocessing module is used to collect multi-source data on contracts, equipment, and finance based on the probe's primary key index structure, and to preprocess the collected data to generate standard data images and abnormal event records. The module is used to perform data normalization processing based on standard data mirroring, identify cost occurrence points, and build a dynamic cost transmission path network by combining predefined allocation rules and triggering conditions. Costs are collected along the path network to generate project-level cost summary data. The synthesis module is used to bind project-level cost summary data with external scenario parameters to generate multiple calculation scenarios. For each scenario, full life cycle economic accounting is performed to obtain a set of economic indicators. Multi-objective optimization analysis is performed on the set of economic indicators, and decision support information is synthesized to generate a structured decision package.

[0095] This embodiment also provides an electronic device suitable for the detailed calculation of the entire life cycle cost of a green electricity direct connection project, including: 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 realize the method for detailed calculation of the entire life cycle cost of a green electricity direct connection project as proposed in the above embodiment.

[0096] This embodiment also provides a storage medium storing a computer program that, when executed by a processor, implements the method for detailed calculation of the full life-cycle cost of green electricity direct connection projects as proposed in the above embodiments.

[0097] The storage medium proposed in this embodiment and the method for detailed calculation of the full life cycle cost of green electricity direct connection projects 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.

[0098] 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.

[0099] 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 method for precise calculation of the full life cycle cost of a green electricity direct connection project, characterized in that: include, Based on the project's basic information and investment and construction model, the main body identification and timeline are standardized, and the multi-source data acquisition probes are deployed and indexed to generate a probe primary key index structure. Based on the probe primary key index structure, multi-source data on contracts, equipment, and finance are collected, and the collected data is preprocessed to generate standard data mirrors and abnormal event records. Based on the standard data mirror, data normalization processing is performed to identify cost occurrence points. Combined with predefined allocation rules and triggering conditions, a dynamic cost transmission path network is constructed. Costs are collected along the path network to generate project-level cost summary data. The project-level cost summary data is bound to external scenario parameters to generate multiple calculation scenarios. Full life-cycle economic accounting is performed on each scenario to obtain a set of economic indicators. Multi-objective optimization analysis is performed on the set of economic indicators, and decision support information is synthesized to generate a structured decision package.

2. The method for detailed calculation of the full life cycle cost of a green electricity direct connection project as described in claim 1, characterized in that: The deployment of the multi-source data acquisition probes includes, Deploy contract-transaction probes for collecting contract execution and transaction settlement data, equipment-maintenance probes for collecting equipment operation and maintenance data, and financial-invoice probes for collecting financial voucher and invoice data.

3. The method for detailed calculation of the full life cycle cost of a green electricity direct connection project as described in claim 2, characterized in that: The points where the identification costs occur include, Identify settlement, default, or price adjustment entries in the contract execution record as contract-related cost incurrence points; Identify billing entries related to capacity, electricity consumption, or ancillary services in grid-connected metering records as transaction cost incurrence points; Replacement, repair, or decommissioning / reactivation entries in equipment operation and maintenance records are identified as points where operation and maintenance costs occur.

4. The method for detailed calculation of the full life cycle cost of a green electricity direct connection project as described in claim 1, characterized in that: The dynamic cost transmission path network is a directed graph structure; The nodes of the directed graph structure are defined by the main body, cost unit, and time point. The edges of the directed graph structure are defined by the allocation rules and triggering conditions, which are used to represent the direction of cost transmission and allocation.

5. The method for detailed calculation of the full life cycle cost of a green electricity direct connection project as described in claim 4, characterized in that: When constructing the dynamic cost transmission path network, it also includes, Read the generated exception event record; Set segment nodes on the path segments covered by the abnormal event records; The soft boundary information and suggested processing path are marked in the edge attributes associated with the segmented nodes.

6. The method for detailed calculation of the full life cycle cost of a green electricity direct connection project as described in claim 1, characterized in that: The multi-objective optimization analysis of the economic indicator set includes, Robust sorting is performed on the set of economic indicators to reduce the impact of missing data and jumps on the sorting results; The ranking results are subjected to Pareto screening to identify non-dominated solution sets as candidate target scenarios.

7. A method for precise calculation of the full life cycle cost of a green electricity direct connection project as described in any one of claims 1-6, characterized in that: The process involves collecting multi-source data on contracts, equipment, and finances based on the probe's primary key index structure, and then preprocessing the collected data. include, Based on the acquisition channels and triggering conditions defined in the probe primary key index structure, contract execution records, equipment operation and maintenance records, and financial invoice records are acquired in parallel. The collected multi-source data undergoes source signature verification and clock alignment processing. Perform missing test labeling and abnormal segment identification on the aligned and checked data; Based on the identification results, a standard data image containing cleaned data is generated, as well as an anomaly event log that records anomaly information and processing suggestions.

8. A system for precise calculation of the full life-cycle cost of green electricity direct-connection projects, using the method described in any one of claims 1-7, characterized in that, include: The generation module is used to standardize the main body identification and timeline based on the project's basic information and investment and construction mode, as well as to deploy and bind the multi-source data acquisition probes to generate the probe primary key index structure. The preprocessing module is used to collect multi-source data on contracts, equipment, and finance based on the probe primary key index structure, and to preprocess the collected data to generate standard data images and abnormal event records. The construction module is used to perform data normalization processing based on the standard data mirror, identify cost occurrence points, and construct a dynamic cost transmission path network in combination with predefined allocation rules and triggering conditions. Costs are collected along the path network to generate project-level cost summary data. The synthesis module is used to bind the project-level cost summary data with external scenario parameters to generate multiple calculation scenarios, perform full life cycle economic accounting for each scenario to obtain a set of economic indicators, perform multi-objective optimization analysis on the set of economic indicators, and synthesize decision support information to generate a structured decision package.

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