A new energy project data processing method, system and device

Abstract

The embodiment of the application provides a new energy project data processing method, system and equipment, the method comprises the following steps: according to the format type of the original review material of the new energy project, the original review material of the new energy project is parsed, and target material text is obtained;The target material text is extracted in direction, and the target index parameter is obtained;According to the task element information, the target standard structured data is obtained from the preset hierarchical rule library;The pre-examination rule in the target standard structured data is called, the target material text is pre-examined, and the pre-examination result is obtained;When the pre-examination result is qualified, the real examination rule in the target standard structured data is called, the target material text is examined, and the real examination result is obtained;When the real examination result is qualified, the comparison rule in the target standard structured data is called, the target index parameter is compared, and the comparison result is obtained.The application can meet the audit demand of project archiving and post-checking.

Description

Technical Field

[0001] The embodiments of the present invention relate to the field of data processing technology, and in particular to a method, system and equipment for processing data of new energy projects. Background Technology

[0002] New energy projects such as wind power, photovoltaics, and energy storage involve large investments and complex technology chains. Feasibility study reports are the core basis for project investment decisions and industry approvals. A complete review requires simultaneous verification of multiple dimensions, including the completeness of materials, the standardization of documents, the suitability of equipment solutions, engineering costs, financial returns, and project risks. The current industry review model has significant shortcomings: First, manual review is time-consuming and labor-intensive, and the review standards are inconsistent. Review experts need to manually review multiple heterogeneous documents, manually extract core indicators such as investment per kilowatt, cost per kilowatt-hour, and internal rate of return, and then manually compare them with industry benchmarks, which is a long process. Review conclusions rely heavily on the experts' professional experience, and different personnel have different review standards, resulting in poor consistency in the review results of similar projects.

[0003] Second, the existing intelligent review system lacks adaptability. Traditional platforms only use static fixed rules, simple keyword matching, or single template verification, without hierarchical management of review rules, and cannot distinguish between the two major scenarios of pre-review material verification and in-depth technical and economic review. When faced with multiple project types such as wind power, photovoltaic, and energy storage, as well as multiple disciplines such as resources, civil engineering, and electrical engineering, they cannot dynamically match exclusive review rules, resulting in frequent problems of rule redundancy, missed detections, and false detections.

[0004] Third, indicator extraction and risk identification are highly dependent on manual labor. Existing tools cannot automatically extract structured economic and technical data across chapters and tables, lack a standardized industry benchmark library for automatic benchmarking, and deviations in cost and revenue indicators can only be investigated manually after the fact, resulting in a lag in the identification of project technical and investment risks.

[0005] Fourth, the review conclusions lack a complete chain of evidence for tracing the source. Traditional manual review opinions are free texts, and each judgment cannot be linked to corresponding review rules, original data coordinates, or benchmarks. Review and tracing require reading the entire report, resulting in weak auditability of the review results and making it difficult to meet the needs of archiving and post-event verification.

[0006] Fifth, the scope of human-machine collaborative review is limited.

[0007] In summary, the existing system can only perform simple numerical logic verification and cannot combine with large models to complete complex professional semantic judgments such as technical routes and risk analysis. It is powerless to conduct in-depth reviews of the rationality of solutions, and the coverage of automated reviews is narrow. Summary of the Invention

[0008] The technical problem to be solved by the embodiments of the present invention is to provide a method, system and equipment for processing new energy project data, which can improve the coverage of automated review by combining rule-based verification.

[0009] To solve the above-mentioned technical problems, the technical solution of the present invention is as follows: A method for processing data from new energy projects includes: Obtain original review materials and task information for new energy projects; Based on the format type of the original review materials for the new energy project, the original review materials for the new energy project are parsed to obtain the target material text; The target material text is extracted in a targeted manner to obtain the target index parameters; Based on the task metadata, target standard structured data is loaded from a preset hierarchical rule base; The pre-screening rules in the target standard structured data are invoked to pre-screen the target material text, and the pre-screening result is obtained; When the preliminary review result is qualified, the actual review rules in the target standard structured data are invoked to conduct an actual review of the target material text and obtain the actual review result. When the audit result is qualified, the comparison rules in the target standard structured data are called to compare the target indicator parameters and obtain the comparison result; The pre-review results, sub-review results, and comparison results are integrated to obtain and output the review results. The construction process of the preset hierarchical rule base includes: obtaining the rule text and tag dimension data of new energy projects; parsing and encapsulating the rule text and tag dimension data of new energy projects to obtain standard structured data; and storing the standard structured data in layers according to the applicable project type tags of the standard structured data to obtain the preset hierarchical rule base.

[0010] Optionally, the original review materials for the new energy project include: text format documents, scanned reports, and table format documents; based on the format type of the original review materials for the new energy project, the original review materials for the new energy project are parsed to obtain the target material text, including: according to To obtain the target material text; in, T For the target material text, Document word It is a text format document. Document pdf This is a scanned report. Document excel It is a table-formatted document. F word This is a function for parsing text-formatted documents.F pdf This is a function for parsing scanned reports. F excel This is a function for parsing table-formatted documents.

[0011] Optionally, the target material text is extracted in a targeted manner to obtain target index parameters, including: according to , thus obtaining the target index parameters; in, P For target indicator parameters, P ={ p 1, p 2, ..., p m}, j = 1 , 2 ,..., m , m The total number of target indicator parameters. p j For a single indicator, T For the target material text, Template index For indicator semantic template library, F extract This is a function for index-oriented extraction.

[0012] Optionally, based on the task metadata, target standard structured data is loaded from a preset hierarchical rule base, including: according to This yields a subset of the target rules. in, M layer Hierarchical labels for target rules in the task metadata; according to The target standard structured data is loaded from the target rule subset; in, S match For target standard structured data, S match ={ ID match , Name match , Logic match , LEVEL match , Prompt match , Tag match}, IDmatch The target is assigned a globally unique identifier number. Name match The name of the target standardization rule. Logic match For the target rule determination logic, LEVEL match To classify and identify target defects. Prompt match The review and rectification opinions were aimed at the target. Tag match For target label dimension data, M type The target project type label in the task metadata. M prof The target professional tag in the task metadata. M stage This refers to the target review stage label in the task metadata. L typei For project type tags, L profi As a professional label, L stagei This is a label for the review stage.

[0013] Optionally, the pre-screening rules in the target standard structured data are invoked to pre-screen the target material text, obtaining pre-screening results, including: according to The pre-screening rules were obtained; in, S pre For the pre-screening rules, S match For target standard structured data, L stagematch The labels for the verification and review phase in the target standard structured data. L layermatch For the verification item type label in the target standard structured data; according to The preliminary review results were obtained; in, Prompt pre For the preliminary review and rectification opinions, LEVEL pre For the classification and identification of defects in the preliminary review, Logic pre For the pre-screening rule determination logic, F pre This is a pre-screening verification function. T The target material text.

[0014] Optionally, the substantive review rules in the target standard structured data are invoked to conduct a substantive review of the target material text, and the review results are obtained, including: according to The substantive review rules were obtained; in, S real For substantive review rules, S match For target standard structured data, L stagematch The labels for the verification and review phase in the target standard structured data. L layermatch For the verification item type label in the target standard structured data; according to , The results of the substantive investigation were obtained; in, Prompt real As the substantive review and rectification opinions, LEVEL real For the classification and identification of defects in substantive audits, Logic real The logic for determining substantive review rules, F real This is the verification function for actual auditing. T The target material text.

[0015] Optionally, a benchmarking analysis is performed on the target indicator parameters to obtain comparison results, including: according to The index deviation rate is obtained. in, Device j The deviation rate of the indicator. p j For a single indicator of the target indicator parameter, j = 1 , 2 ,..., m , m The total number of target indicator parameters. B j This serves as a benchmark value. according to The comparison results were obtained. in, Risk j For comparison results, Device j The deviation rate of the indicator. j = 1 , 2 ,..., m , mThe total number of target indicator parameters. Threshold low The first threshold, Threshold high This is the second threshold.

[0016] Optionally, the rule text and tag dimension data of the new energy project are parsed and encapsulated to obtain standard structured data, including: The rule text and tag dimension data of the new energy project are paired to obtain a binary input group; The binary input group is parsed and processed to obtain a structured dataset; The structured dataset is encapsulated to obtain standard structured data.

[0017] Embodiments of the present invention also provide a data processing system for new energy projects, comprising: The acquisition module is used to acquire the original review materials and task element information of new energy projects; The processing module is used to parse the original review materials of the new energy project according to their format type to obtain target material text; to extract target indicator parameters from the target material text; to load target standard structured data from a preset hierarchical rule base according to the task element information; to call the pre-review rules in the target standard structured data to pre-review the target material text and obtain a pre-review result; and when the pre-review result is qualified, to call the actual review rules in the target standard structured data to conduct an actual review of the target material text and obtain an actual review result. When the audit result is qualified, the comparison rules in the target standard structured data are invoked to compare the target indicator parameters and obtain the comparison result; the pre-audit result, audit result and comparison result are integrated to obtain and output the review result; wherein, the construction process of the preset hierarchical rule library includes: obtaining the rule text and tag dimension data of new energy projects; parsing and encapsulating the rule text and tag dimension data of new energy projects to obtain standard structured data; according to the applicable project type tags of the standard structured data, storing the standard structured data in layers to obtain the preset hierarchical rule library.

[0018] Embodiments of the present invention also provide a computing device, comprising: One or more processors; A storage device for storing one or more programs that, when executed by one or more processors, cause the one or more processors to perform the method as described above.

[0019] The above-described solutions of the embodiments of the present invention have at least the following beneficial effects: The above-described solution in this invention, through automated material analysis, indicator extraction, and hierarchical rule verification, replaces manual review and benchmarking, significantly shortening the review cycle, avoiding review biases caused by differences in human experience, and ensuring consistent and standardized review results for similar projects. A hierarchical rule base is constructed to separate pre-review and actual review scenarios, dynamically matching specific review rules for different new energy projects, eliminating the redundancy, omissions, and errors in traditional system rules, and improving system adaptability and review accuracy. Simultaneously, fully automated indicator capture and intelligent benchmarking are achieved, proactively identifying deviations in cost, revenue, and other indicators, as well as project risks, solving the problem of delayed risk identification. Finally, multi-dimensional review results are integrated to form a complete review basis, establishing a traceable review evidence chain, and rule-based verification enhances the coverage of automated review, meeting the audit requirements for project archiving and post-audit verification. Attached Figure Description

[0020] Figure 1 This is a flowchart illustrating the data processing method for new energy projects provided in an embodiment of the present invention.

[0021] Figure 2 This is a schematic diagram of the modules of the new energy project data processing system provided in the embodiments of the present invention. Detailed Implementation

[0022] Exemplary embodiments of the invention will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the invention are shown in the drawings, it should be understood that the invention may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this invention will be thorough and complete, and will fully convey the scope of the invention to those skilled in the art.

[0023] like Figure 1 As shown, an embodiment of the present invention provides a method for processing data from new energy projects, including: Step 11: Obtain the original review materials and task element information for the new energy project; Step 12: Based on the format type of the original review materials for the new energy project, parse the original review materials for the new energy project to obtain the target material text; Step 13: Targeted extraction of the target material text to obtain target index parameters; Step 14: Based on the task metadata, load the target standard structured data from the preset hierarchical rule base; Step 15: Call the pre-screening rules in the target standard structured data to pre-screen the target material text and obtain the pre-screening result; Step 16: When the preliminary review result is qualified, the actual review rules in the target standard structured data are called to conduct an actual review of the target material text and obtain the actual review result; Step 17: When the audit result is qualified, the comparison rules in the target standard structured data are called to compare the target indicator parameters and obtain the comparison result; Step 18: Integrate the preliminary review results, substantive review results, and comparison results to obtain and output the review results; The construction process of the preset hierarchical rule base includes: Step 101: Obtain the rule text and tag dimension data for new energy projects; Step 102: Parse and encapsulate the new energy project rule text and tag dimension data to obtain standard structured data; Step 103: Based on the applicable project type tags of the standard structured data, store the standard structured data in layers to obtain a preset layered rule library.

[0024] In this embodiment, automated material analysis, indicator extraction, and hierarchical rule verification replace manual review and benchmarking, significantly shortening the review cycle, avoiding review biases caused by differences in human experience, and ensuring consistent and standardized review results for similar projects. A hierarchical rule base is constructed to separate pre-review and actual review scenarios, dynamically matching specific review rules for different new energy projects, eliminating the redundancy, omissions, and errors in traditional system rules, and improving system adaptability and review accuracy. Simultaneously, fully automated indicator capture and intelligent benchmarking are achieved, proactively identifying deviations in cost, revenue, and other indicators, as well as project risks, resolving the problem of delayed risk identification. Finally, multi-dimensional review results are integrated to form a complete review basis, establishing a traceable review evidence chain, and rule-based verification enhances the coverage of automated reviews, meeting the audit requirements for project archiving and post-audit verification.

[0025] In an optional embodiment of the present invention, step 101, obtaining the new energy project rule text and tag dimension data, includes: Step 1011: Collect the original rule texts for the review of new energy projects from multiple channels; obtain national industry standards and specifications, enterprise internal review management systems, historical project expert review experience, and technical and economic special review constraint documents, and summarize these documents to obtain the original rule texts for the review of new energy projects. Step 1012: Collect tag dimension data including project type tags (values ​​include photovoltaic, wind power, energy storage, and wind-solar-storage composite projects), professional tags (values ​​include resource professional, civil engineering professional, primary electrical, secondary electrical, economic evaluation, and general planning), review stage tags (values ​​include pre-review and substantive review), and rule layer tags (values ​​include first-level identifier and second-level identifier; the first-level identifier corresponds to pre-review rules for material completeness and format compliance, and the second-level identifier corresponds to substantive review rules for technical solutions and economic indicators).

[0026] This embodiment, through the collection of original rule texts for new energy project review and the collection of tag dimension data, unifies and aggregates scattered review clauses from different documents, systems, and experts into a centralized original rule text pool. Subsequent batch text cleaning, semantic parsing, and logical transformation processing can be performed without repeatedly retrieving rule content across documents, significantly reducing data processing costs during the rule base construction phase. This provides a complete and unified foundation of original materials for subsequent binary input pairing and structured encapsulation. It covers the entire professional chain of resources, civil engineering, primary and secondary electrical systems, economic evaluation, and general planning, supporting parallel reviews by multiple disciplines. The system can individually retrieve the corresponding professional verification rules based on the professional tags of the review task, enabling module-based targeted review. This addresses the shortcomings of traditional reviews that lack professional segmentation and suffer from ambiguous problem identification, matching the business scenario of multi-professional collaborative review for new energy projects.

[0027] In an optional embodiment of the present invention, in step 102, the new energy project rule text and tag dimension data are parsed and encapsulated to obtain standard structured data, including: Step 1021: Pair the new energy project rule text and tag dimension data to obtain a binary input group; specifically, according to This yields a binary input set; in, Input i For a binary input group, r i For the rules and regulations of new energy projects, Tag i For label dimension data, i=1 , 2 ,..., n , n This represents the total number of rule texts for new energy projects. Tag i = { L typei , L profi , L stagei , Llayeri}, L typei For project type tags, L profi As a professional label, L stagei For the review stage label, L layeri Hierarchical labels for rules; Step 1022: The binary input group is parsed and processed to obtain a structured dataset; specifically, according to... This yields a structured dataset; in, s i For structured datasets, s i ={ ID i , Name i , Logic i , LEVEL i , Prompt i , Tag i}, ID i It is a globally unique identifier. Name i For the standardized rule name, Logic i This is the rule-based decision logic (which transforms natural language rules into machine-recognizable judgment conditions). LEVEL i For defect classification identification, Prompt i In order to review the rectification opinions, Tag i For label dimension data, Input i For a binary input group, r i For the rules and regulations of new energy projects, i=1 , 2 ,..., n , n This represents the total number of rule texts for new energy projects. F jx To analyze the processing function, F bh For number generation function, F mc For name standardization processing functions, F lj This is a logic parsing conversion function (natural language logic to machine decision logic conversion).F qx This is the defect level determination function. F ps Provide matching functions for the review; Step 1023: Encapsulate the structured dataset to obtain standard structured data; Specifically, according to This yields standard structured data; in, S For standard structured data, s i For structured datasets, i=1 , 2 ,..., n , n This represents the total number of rule texts for new energy projects. F map This is a mapping wrapper function.

[0028] In this embodiment, the original text rules are bound one-to-one with four-dimensional tags to form binary groups, achieving a strong association between rules and applicable scenarios. Through various processing functions, natural language clauses are automatically converted into structured units with unique numbers, standardized names, machine-executable judgment logic, defect levels, and rectification statements, eliminating problems of messy text descriptions and inconsistent standards. Batch encapsulation forms a unified dataset, transforming scattered expert experience and industry standards into standardized knowledge assets that can be batch-searched and automatically scheduled by machines. This supports the system in accurately filtering and hierarchically calling rules, solving the pain point that traditional plain text rules cannot be automatically verified. It provides a standardized data foundation for subsequent dynamic routing and tiered review, significantly improving the automation level and execution stability of intelligent review.

[0029] In an optional embodiment of the present invention, step 103 involves storing the standard structured data in layers according to the rule-based layering tags of the standard structured data to obtain a preset layered rule library, including: Step 1031, according to This yields two sets of rule subsets; in, S 1 represents the first set of rules (corresponding to the integrity, format, and signature compliance rules used in the pre-review stage). S 2 is the second set of rules (corresponding to the in-depth review rules for technical solutions, indicators, and risk analysis used in the substantive review stage). s i For structured datasets, i=1 , 2 ,..., n , n This represents the total number of rule texts for new energy projects. S For standard structured data, Llayeri Hierarchical labels for rules; Step 1032: Partition and store the two sets of rule subsets to obtain a preset hierarchical rule base. Database rule The two types of rules are stored in separate areas of the database and a fast retrieval index is created to form a hierarchical rule database. When conducting project reviews, the system can retrieve the rules in the corresponding partitions according to the review stage and level requirements, so as to realize the on-demand loading of rules. At the same time, it supports the addition, modification, and version management of rules to adapt to business standard updates.

[0030] In this embodiment, the rules are automatically split into two subsets—preliminary review and substantive review—based on hierarchical tags, enabling physical partitioning and corresponding retrieval indexes for both types of review logic. During review, the corresponding partition rules can be retrieved individually as needed, eliminating the need to load all rules at once, reducing system computational redundancy, and improving rule matching and scheduling speed. It distinguishes between shallow material verification and in-depth technical and economic validation logic, enabling phased, progressive review, early interception of format defects, and avoidance of ineffective in-depth reviews. Simultaneously, the database supports rule addition, modification, and version control. When industry standards and review requirements are updated, the entire rule system does not need to be reconstructed, adapting to business iterations, significantly reducing rule maintenance costs, and ensuring the long-term scalability and stable operation of the rule base.

[0031] In an optional embodiment of the present invention, in step 11, the original review materials and task element information of the new energy project are obtained. Specifically, all the project's application documents are first collected and the version information is retained. Then, the project, professional, and stage attribute tags of this review are entered. The two types of data are bound and archived as a unified input source for subsequent document parsing, rule matching, and intelligent review, providing basic materials and scenario matching basis for the full-process hierarchical automatic review.

[0032] In this embodiment, all types of application documents are uniformly collected and versions are retained to avoid confusion between multiple versions of documents. Meta tags such as project type, specialty, and review stage are simultaneously entered and bound to the materials for archiving, forming a complete data source for review tasks. This provides accurate scenario basis for subsequent document parsing and intelligent rule matching. The system can automatically filter and adapt review rules based on meta information, eliminating the need for manual differentiation of project categories and review scopes. This ensures a smooth hierarchical review process from the source, reduces the workload of manual information entry and filtering, and improves the overall efficiency of automated review.

[0033] In an optional embodiment of the present invention, in step 12, the original review materials for the new energy project include: text format documents, scanned reports, and table format documents; according to the format type of the original review materials for the new energy project, the original review materials for the new energy project are parsed to obtain the target material text, including: Step 121, according to To obtain the target material text; in, T For the target material text, Document word It is a text format document. Document pdf This is a scanned report. Document excel It is a table-formatted document. F word This is a function for parsing text-formatted documents. F pdf This is a function for parsing scanned reports. F excel This is a function for parsing table-formatted documents.

[0034] In this embodiment, dedicated parsing functions are configured for three types of differentiated application documents to uniformly output standardized structured text, solving the problem of inconsistent recognition of heterogeneous documents. It can automatically extract text content from the main body, tables, and drawings, retaining text position anchors and eliminating the need for manual transcription and formatting. The unified data format provides standardized input for subsequent indicator extraction and rule verification, avoiding recognition deviations caused by different file formats, and comprehensively improving the automation level of document processing and the accuracy of data extraction.

[0035] In an optional embodiment of the present invention, step 13 involves targeted extraction of the target material text to obtain target index parameters, including: Step 131, according to , thus obtaining the target index parameters; in, P For target indicator parameters, P ={ p 1, p 2, ..., p m}, j = 1 , 2 ,..., m , m The total number of target indicator parameters. p j For a single indicator, T For the target material text, Template index It serves as a semantic template library for indicators (the library stores various new energy indicator standard names, multiple synonyms, numerical matching regular expressions, and unified unit conversion rules, covering core review indicators such as power generation, investment per kilowatt, cost per kilowatt, internal rate of return, and net present value). F extract This is a function for index-oriented extraction.

[0036] In this embodiment, relying on a pre-built indicator template library, extraction functions are used to automatically extract various indicator data from standardized text. This approach is compatible with multiple synonyms for indicators and uses standardized units, eliminating the need for manual table-by-table copying and significantly reducing labor costs. It can fully cover core review parameters such as power generation, cost, and revenue, synchronously binding the original text location of indicators to avoid errors and omissions from manual copying. The output is a well-organized set of indicators, providing accurate and unified data input for subsequent indicator deviation calculations and risk classification comparisons, effectively improving indicator extraction efficiency and data reliability.

[0037] In an optional embodiment of the present invention, step 14, loading target standard structured data from a preset hierarchical rule base according to the task metadata, includes: Step 141, according to This yields a subset of the target rules. in, M layer Hierarchical labels for target rules in the task metadata; Step 142, according to The target standard structured data is loaded from the target rule subset; in, S match For target standard structured data, S match ={ ID match , Name match , Logic match , LEVEL match , Prompt match , Tag match}, ID match The target is assigned a globally unique identifier number. Name match The name of the target standardization rule. Logic match For the target rule determination logic, LEVEL match To classify and identify target defects. Prompt match The review and rectification opinions were aimed at the target. Tag match For target label dimension data, M type The target project type label in the task metadata. M prof The target professional tag in the task metadata.M stage This refers to the target review stage label in the task metadata. L typei For project type tags, L profi As a professional label, L stagei This is a label for the review stage.

[0038] In this embodiment, the rule subsets corresponding to the pre-review or sub-review are first identified based on the task layering identifier. Then, through multi-dimensional matching logic, structured rules that perfectly match the current project type, review specialty, and review stage are accurately selected, while irrelevant rules are filtered out to avoid resource consumption and misjudgments caused by full loading. A dedicated rule set carrying complete judgment logic, defect levels, and rectification suggestions is output, providing accurate basis for subsequent automatic verification. No manual screening or differentiation of applicable clauses is required, adapting to various types of new energy projects and multi-specialty parallel review scenarios, improving rule matching accuracy and overall review efficiency.

[0039] In an optional embodiment of the present invention, step 15 involves calling the pre-screening rules in the target standard structured data to pre-screen the target material text and obtain a pre-screening result, including: Step 151, according to The pre-screening rules were obtained; in, S pre For the pre-screening rules, S match For target standard structured data, L stagematch The labels for the verification and review phase in the target standard structured data. L layermatch For the verification item type label in the target standard structured data; Step 152, according to The preliminary review results were obtained; in, Prompt pre For the preliminary review and rectification opinions, LEVEL pre For the classification and identification of defects in the preliminary review, Logic pre For the pre-screening rule determination logic, F pre For pre-screening verification functions (used for line-by-line comparison of structured target material text) TThe pre-review rules are logically validated one by one. During the validation process, the unique rule number corresponding to each defect, the anchor point of the defect content in the original document, and the matching standardized rectification script are automatically recorded. If any serious defect is detected during the validation process, the pre-review is directly deemed unqualified and the overall review process is terminated. If there are no serious defects, the pre-review is deemed qualified and the process is transferred to the subsequent actual review stage. T The target material text; Specifically, the preliminary review results are obtained according to the preset grading criteria: when the output of the preliminary review verification function combined with the target material text and the preliminary review rules is 1, the preliminary review rectification opinion is marked as qualified, and the preliminary review defect grading is marked as no serious defect; when the calculation result is 0.9, the preliminary review rectification opinion is missing attachments, and the preliminary review defect grading is marked as a general defect; when the calculation result is 0.8, the preliminary review rectification opinion is missing attached figures, and the preliminary review defect grading is marked as a general defect; when the calculation result is 0.7, the preliminary review rectification opinion is missing feasibility study report, and the preliminary review defect grading is marked as a major defect; when the calculation result is 0.6, the preliminary review rectification opinion is missing resource review report, and the preliminary review defect grading is marked as a major defect; when the calculation result is 0.5, the preliminary review rectification opinion is unsuccessful signature recognition, and the preliminary review defect grading is marked as a major defect; except for the scenarios corresponding to the above values, for all other calculation result scenarios, the preliminary review rectification opinion is uniformly classified as other situations, and the preliminary review defect grading is marked as a serious defect.

[0040] In this embodiment, the first-level pre-review rules are precisely selected, and only basic issues such as material completeness and format compliance are checked. The pre-review verification function automatically matches each document text, and simultaneously records the rule number, original text location, and rectification opinions corresponding to the defects. Once a serious material defect is detected, the process is terminated directly, avoiding invalid entry into the in-depth review, which greatly saves system computing power and review time. Low-level material problems are intercepted in advance, reducing repeated returns for modification later, while retaining complete defect traceability information, realizing standardized automatic control of pre-review and unifying the review standards for basic materials.

[0041] In an optional embodiment of the present invention, step 16 involves invoking the substantive review rules in the target standard structured data to conduct a substantive review of the target material text and obtain the substantive review result, including: Step 161, according to The substantive review rules were obtained; in, S real For substantive review rules, S match For target standard structured data, L stagematch The labels for the verification and review phase in the target standard structured data.L layermatch For the verification item type label in the target standard structured data; Step 162, according to , The results of the substantive investigation were obtained; in, Prompt real As the substantive review and rectification opinions, LEVEL real For the classification and identification of defects in substantive audits, Logic real The logic for determining substantive review rules, F real The audit verification function (this function relies on the structured target material text to simultaneously complete three types of processing: First, it performs procedural rigid numerical verification to check quantifiable rules such as parameter consistency and chapter completeness; Second, it uses the Model Context Protocol (MCP) to encapsulate the calling interfaces of the underlying calculation tools, benchmark databases, and enterprise internal cloud platforms, and uses standardized APIs to call the enterprise's internal economic evaluation calculation tools to obtain authoritative financial calculation indicators such as Internal Rate of Return (IRR), Net Present Value (NPV), and Levelized Cost of Electricity (LCOE); Third, it links with the large language model to conduct in-depth semantic analysis of the technical solution. In the entire verification process, each identified defect is bound to a unique rule number, the anchor point of the indicator or text in the original document, and standardized professional rectification suggestions; if any serious technical or economic flaw is detected during the verification process, the audit is directly deemed unqualified and the overall review process is terminated; if no serious defects are detected, the audit is deemed qualified and the process proceeds to the subsequent indicator comparison stage). T The target material text; Specifically, the audit results are obtained according to the preset grading criteria: when the audit verification function, combined with the target material text and the audit rules, outputs a result of 1, the audit review rectification opinion is marked as qualified, and the audit defect grading is marked as no serious defect; when the calculation result is 0.5, the audit review rectification opinion is that the document is missing a chapter, and the audit defect grading is marked as a major defect; when the calculation result is 0.4, the audit review rectification opinion is that the document is missing a solution, and the audit defect grading is marked as a major defect; when the calculation result is 0.3, the audit review rectification opinion is that the document solution has a rationality problem, and the audit defect grading is marked as a major defect; when the calculation result is 0.2, the audit review rectification opinion is that the document solution has an economic problem, and the audit defect grading is marked as a major defect; except for the scenarios corresponding to the above values, for all other calculation result scenarios, the audit review rectification opinion is uniformly classified as other situations, and the audit defect grading is marked as a serious defect.

[0042] In this embodiment, the second-layer in-depth review rules are precisely selected, focusing on technical solutions and economic indicators for professional review. The review verification function integrates numerical rigidity verification and large-scale model semantic analysis, taking into account both data accuracy and the rationality of the solution. Each defect is bound to a rule number, original text anchor point, and professional rectification suggestions. The detection of major technical and economic flaws can directly terminate the process, avoiding invalid indicator benchmarking. This achieves automation of multi-professional in-depth review, unifies technical and economic review standards, reduces expert subjective bias, and retains complete defect traceability information, improving the professionalism and traceability of the review.

[0043] In an optional embodiment of the present invention, step 17, performing an index benchmarking analysis on the target index parameters to obtain comparison results, includes: Step 171, according to The index deviation rate is obtained. in, Device j The deviation rate of the indicator. p j For a single indicator of the target indicator parameter, j = 1 , 2 ,..., m , m The total number of target indicator parameters. B j This serves as a benchmark value. Step 172, according to The comparison results were obtained. in, Risk j For comparison results, Device j The deviation rate of the indicator. j = 1 , 2 ,..., m , m The total number of target indicator parameters. Threshold low The first threshold, Threshold high This is the second threshold.

[0044] In this embodiment, the deviation rate of each indicator relative to the benchmark value is automatically calculated using a formula. Based on a two-level threshold, the indicators are divided into three risk levels: normal, watch out, and warning. This enables quantitative risk identification of investment and return indicators without the need for manual comparison and calculation. The quantitative deviation directly reflects the degree of deviation of the project's economic indicators, automatically identifying high-risk indicators and exposing potential investment risks in advance. A unified risk classification standard eliminates the subjectivity of manual judgment, while generating standardized indicator comparison results, providing quantitative data support for the final review conclusion and improving the objectivity and efficiency of project economic risk review.

[0045] In an optional embodiment of the present invention, step 18 integrates the preliminary review results, substantive review results, and comparison results to obtain the review results, including: Step 181, The review results were obtained. in, Algorithm Structure For the review results, Prompt pre For the preliminary review and rectification opinions, LEVEL pre For the classification and identification of defects in the preliminary review, Prompt real As the substantive review and rectification opinions, LEVEL real For the classification and identification of defects in substantive audits, Risk j For comparison results.

[0046] In this embodiment, three types of comprehensive review data—preliminary review, substantive review, and indicator benchmarking—are collected and uniformly packaged into a complete and standardized review result. This integrates all review information related to materials, technology, and economic risks, avoiding the fragmentation and difficulty in summarizing and viewing various review results. The review result simultaneously carries the defect level, rectification suggestions, and indicator risk classification for each stage, fully covering the review conclusions of the project from all dimensions. This eliminates the need for manual compilation of multiple review records, facilitating experts to quickly and comprehensively assess the project's strengths and weaknesses. Simultaneously, it completely retains the review data throughout the entire process, supporting report output, post-review verification, and archiving, thereby improving the completeness and reusability of the review results.

[0047] The above-mentioned solution of this invention collects review rules through multiple channels and builds a four-dimensional label system. After parsing, encapsulation, and hierarchical storage, it constructs an iterative hierarchical rule library and uniformly precipitates standardized review knowledge. The review end collects multi-format application materials and task meta-information, and extracts indicators through heterogeneous document parsing and intelligent extraction. Based on the meta-information, it accurately matches the corresponding pre-review / substantive review rules, and verifies in stages. First, it intercepts low-level defects in the materials, then integrates numerical verification and large-scale model analysis to conduct in-depth technical review, and then uses benchmark formulas to quantify and classify the risk level of indicators. Finally, it summarizes the data of the entire process to generate an integrated review result. The entire process is fully automated and standardized, greatly reducing the workload of manual copying, screening, and comparison, and eliminating subjective review bias of experts. Staged verification saves computing power, and the entire process retains rule numbers and original text anchors to form a complete traceability chain. The quantitative classification of indicators can identify potential investment risks in advance. The rule library supports iterative updates and is adapted to the review of multiple types and professional new energy projects such as photovoltaic, wind power, and energy storage, taking into account review efficiency, review objectivity, and auditability of results.

[0048] like Figure 2 As shown, embodiments of the present invention also provide a data processing system 20 for new energy projects, comprising: Module 21 is used to acquire the original review materials and task element information of new energy projects; Processing module 22 is used to parse the original review materials of the new energy project according to the format type to obtain target material text; to extract target indicator parameters from the target material text; to load target standard structured data from a preset hierarchical rule base according to the task element information; to call the pre-review rules in the target standard structured data to pre-review the target material text and obtain a pre-review result; and when the pre-review result is qualified, to call the actual review rules in the target standard structured data to conduct an actual review of the target material text and obtain the actual review result. As a result, when the audit result is qualified, the comparison rules in the target standard structured data are invoked to compare the target indicator parameters and obtain the comparison result; the pre-audit result, audit result and comparison result are integrated to obtain and output the review result; wherein, the construction process of the preset hierarchical rule library includes: obtaining the rule text and tag dimension data of new energy projects; parsing and encapsulating the rule text and tag dimension data of new energy projects to obtain standard structured data; according to the applicable project type tags of the standard structured data, storing the standard structured data in layers to obtain the preset hierarchical rule library.

[0049] Optionally, the original review materials for the new energy project include: text format documents, scanned reports, and table format documents; based on the format type of the original review materials for the new energy project, the original review materials for the new energy project are parsed to obtain the target material text, including: according to To obtain the target material text; in, T For the target material text, Document word It is a text format document. Document pdf This is a scanned report. Document excel It is a table-formatted document. F word This is a function for parsing text-formatted documents. F pdf This is a function for parsing scanned reports. F excel This is a function for parsing table-formatted documents.

[0050] Optionally, the target material text is extracted in a targeted manner to obtain target index parameters, including: according to , thus obtaining the target index parameters; in, P For target indicator parameters,P ={ p 1, p 2, ..., p m}, j = 1 , 2 ,..., m , m The total number of target indicator parameters. p j For a single indicator, T For the target material text, Template index For indicator semantic template library, F extract This is a function for index-oriented extraction.

[0051] Optionally, based on the task metadata, target standard structured data is loaded from a preset hierarchical rule base, including: according to This yields a subset of the target rules. in, M layer Hierarchical labels for target rules in the task metadata; according to The target standard structured data is loaded from the target rule subset; in, S match For target standard structured data, S match ={ ID match , Name match , Logic match , LEVEL match , Prompt match , Tag match}, ID match The target is assigned a globally unique identifier number. Name match The name of the target standardization rule. Logic match For the target rule determination logic, LEVEL match To classify and identify target defects. Prompt match The review and rectification opinions were aimed at the target. Tag match For target label dimension data, Mtype The target project type label in the task metadata. M prof The target professional tag in the task metadata. M stage This refers to the target review stage label in the task metadata. L typei For project type tags, L profi As a professional label, L stagei This is a label for the review stage.

[0052] Optionally, the pre-screening rules in the target standard structured data are invoked to pre-screen the target material text, obtaining pre-screening results, including: according to The pre-screening rules were obtained; in, S pre For the pre-screening rules, S match For target standard structured data, L stagematch The labels for the verification and review phase in the target standard structured data. L layermatch For the verification item type label in the target standard structured data; according to The preliminary review results were obtained; in, Prompt pre For the preliminary review and rectification opinions, LEVEL pre For the classification and identification of defects in the preliminary review, Logic pre For the pre-screening rule determination logic, F pre This is a pre-screening verification function. T The target material text.

[0053] Optionally, the substantive review rules in the target standard structured data are invoked to conduct a substantive review of the target material text, and the review results are obtained, including: according to The substantive review rules were obtained; in, S real For substantive review rules, S match For target standard structured data, L stagematch The labels for the verification and review phase in the target standard structured data. Llayermatch For the verification item type label in the target standard structured data; according to , The results of the substantive investigation were obtained; in, Prompt real As the substantive review and rectification opinions, LEVEL real For the classification and identification of defects in substantive audits, Logic real The logic for determining substantive review rules, F real This is the verification function for actual auditing. T The target material text.

[0054] Optionally, a benchmarking analysis is performed on the target indicator parameters to obtain comparison results, including: according to The index deviation rate is obtained. in, Device j The deviation rate of the indicator. p j For a single indicator of the target indicator parameter, j = 1 , 2 ,..., m , m The total number of target indicator parameters. B j This serves as a benchmark value. according to The comparison results were obtained. in, Risk j For comparison results, Device j The deviation rate of the indicator. j = 1 , 2 ,..., m , m The total number of target indicator parameters. Threshold low The first threshold, Threshold high This is the second threshold.

[0055] Optionally, the rule text and tag dimension data of the new energy project are parsed and encapsulated to obtain standard structured data, including: The rule text and tag dimension data of the new energy project are paired to obtain a binary input group; The binary input group is parsed and processed to obtain a structured dataset; The structured dataset is encapsulated to obtain standard structured data.

[0056] It should be noted that this device is a device corresponding to the above method. All implementation methods in the above method embodiments are applicable to this embodiment and can achieve the same technical effect.

[0057] Embodiments of the present invention also provide a computing device, including: one or more processors; and a storage device for storing one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement the method described above. All implementations in the above method embodiments are applicable to this embodiment and can achieve the same technical effect.

[0058] Embodiments of the present invention also provide a computing device readable storage medium storing instructions that, when executed on a computing device, cause the computing device to perform the method described above. All implementations in the above method embodiments are applicable to this embodiment and can achieve the same technical effect.

[0059] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed in this invention can be implemented in electronic hardware, or a combination of computing device software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0060] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0061] In the embodiments provided by this invention, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative. For instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0062] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0063] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0064] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computing device-readable storage medium. Based on this understanding, the technical solution of this invention, essentially, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computing device software product is stored in a storage medium and includes several instructions to cause a computing device (which may be a personal computing device, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, ROM, RAM, magnetic disks, or optical disks.

[0065] Furthermore, it should be noted that in the apparatus and method of the present invention, it is obvious that the components or steps can be decomposed and / or recombined. These decompositions and / or recombinations should be considered equivalent solutions of the present invention. Moreover, the steps performing the above-described series of processes can naturally be executed in the order described, but are not necessarily required to be executed in chronological order; some steps can be executed in parallel or independently of each other. Those skilled in the art will understand that all or any step or component of the method and apparatus of the present invention can be implemented in any computing device (including processors, storage media, etc.) or network of computing devices, in hardware, firmware, software, or a combination thereof. This is something that those skilled in the art can achieve using basic programming skills after reading the description of the present invention.

[0066] Therefore, the object of the present invention can also be achieved by running a program or a set of programs on any computing device. The computing device can be a known general-purpose device. Therefore, the object of the present invention can also be achieved simply by providing a program product containing program code implementing the method or apparatus. That is, such a program product also constitutes the present invention, and the storage medium storing such a program product also constitutes the present invention. Obviously, the storage medium can be any known storage medium or any storage medium developed in the future. It should also be noted that in the apparatus and method of the present invention, it is obvious that the components or steps can be decomposed and / or recombined. These decompositions and / or recombinations should be considered equivalent to the present invention. Furthermore, the steps performing the above series of processes can naturally be performed in the order described, but are not necessarily required to be performed in chronological order. Some steps can be performed in parallel or independently of each other.

[0067] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for processing data from new energy projects, characterized in that, include: Obtain original review materials and task information for new energy projects; Based on the format type of the original review materials for the new energy project, the original review materials for the new energy project are parsed to obtain the target material text; The target material text is extracted in a targeted manner to obtain the target index parameters; Based on the task metadata, target standard structured data is loaded from a preset hierarchical rule base; The pre-screening rules in the target standard structured data are invoked to pre-screen the target material text, and the pre-screening result is obtained; When the preliminary review result is qualified, the actual review rules in the target standard structured data are invoked to conduct an actual review of the target material text and obtain the actual review result. When the audit result is qualified, the comparison rules in the target standard structured data are called to compare the target indicator parameters and obtain the comparison result; The pre-review results, sub-review results, and comparison results are integrated to obtain and output the review results. The construction process of the preset hierarchical rule base includes: obtaining the rule text and tag dimension data of new energy projects; parsing and encapsulating the rule text and tag dimension data of new energy projects to obtain standard structured data; and storing the standard structured data in layers according to the applicable project type tags of the standard structured data to obtain the preset hierarchical rule base.

2. The method for processing new energy project data according to claim 1, characterized in that, The original review materials for the new energy projects include: text-formatted documents, scanned reports, and table-formatted documents. Based on the format type of the original review materials, the materials are parsed to obtain the target material text, including: according to To obtain the target material text; in, T For the target material text, Doc word It is a text format document. Doc pdf This is a scanned report. Doc excel It is a table-formatted document. F word This is a function for parsing text-formatted documents. F pdf This is a function for parsing scanned reports. F excel This is a function for parsing table-formatted documents.

3. The method for processing new energy project data according to claim 1, characterized in that, The target material text is extracted in a targeted manner to obtain target index parameters, including: according to , thus obtaining the target indicator parameters; in, P For target indicator parameters, P ={ p 1, p 2, ..., p m }, j = 1 , 2 ,..., m , m The total number of target indicator parameters. p j For a single indicator, T For the target material text, Temp index For indicator semantic template library, F extract This is a function for index-oriented extraction.

4. The method for processing new energy project data according to claim 1, characterized in that, Based on the task metadata, target standard structured data is loaded from a preset hierarchical rule base, including: according to This yields a subset of the target rules. in, M layer Hierarchical labels for target rules in the task metadata; according to The target standard structured data is loaded from the target rule subset; in, S match For target standard structured data, S match ={ ID match , Name match , Logic match , LEVEL match , Tip match , Tag match }, ID match The target is assigned a globally unique identifier number. Name match The name of the target standardization rule. Logic match For the target rule determination logic, LEVEL match To classify and identify target defects. Tip match The review and rectification opinions were aimed at the target. Tag match For target label dimension data, M type The target project type label in the task metadata. M prof The target professional tag in the task metadata. M stage This refers to the target review stage label in the task metadata. L typei For project type tags, L profi As a professional label, L stagei This is a label for the review stage.

5. The method for processing new energy project data according to claim 1, characterized in that, The pre-screening rules in the target standard structured data are invoked to pre-screen the target material text, and the pre-screening results are obtained, including: according to The pre-screening rules were obtained. in, S pre For the pre-screening rules, S match For target standard structured data, L stagematch The labels for the verification and review phase in the target standard structured data. L layermatch For the verification item type label in the target standard structured data; according to The preliminary review results were obtained; in, Tip pre For the preliminary review and rectification opinions, LEVEL pre For the classification and identification of defects in the preliminary review, Logic pre For the pre-screening rule determination logic, F pre This is a pre-screening verification function. T The target material text.

6. The method for processing new energy project data according to claim 1, characterized in that, The substantive review rules in the target standard structured data are invoked to conduct a substantive review of the target material text, and the review results are obtained, including: according to The substantive review rules were obtained; in, S real For substantive review rules, S match For target standard structured data, L stagematch The labels for the verification and review phase in the target standard structured data. L layermatch For the verification item type label in the target standard structured data; according to ， The results of the substantive investigation were obtained; in, Tip real As part of the substantive review and rectification opinions, LEVEL real For the classification and identification of defects in substantive audits, Logic real The logic for determining substantive review rules, F real This is the verification function for actual auditing. T The target material text.

7. The method for processing new energy project data according to claim 1, characterized in that, A benchmarking analysis was performed on the target indicator parameters to obtain the comparison results, including: according to The index deviation rate is obtained. in, Dev j The deviation rate of the indicator. p j For a single indicator of the target indicator parameter, j = 1 , 2 ,..., m , m The total number of target indicator parameters. B j This serves as a benchmark value. according to The comparison results were obtained. in, Risk j For comparison results, Dev j The deviation rate of the indicator. j = 1 , 2 ,..., m , m The total number of target indicator parameters. Th low The first threshold, Th high This is the second threshold.

8. The method for processing new energy project data according to claim 1, characterized in that, The rule text and tag dimension data of the new energy projects are parsed and encapsulated to obtain standard structured data, including: The rule text and tag dimension data of the new energy project are paired to obtain a binary input group; The binary input group is parsed and processed to obtain a structured dataset; The structured dataset is encapsulated to obtain standard structured data.

9. A data processing system for new energy projects, characterized in that, include: The acquisition module is used to acquire the original review materials and task element information of new energy projects; The processing module is used to parse the original review materials of the new energy project according to the format type to obtain the target material text; to extract the target indicator parameters from the target material text; and to load the target standard structured data from the preset hierarchical rule base according to the task element information. The process involves: 1) calling the pre-review rules in the target standard structured data to pre-review the target material text and obtain a pre-review result; 2) calling the actual review rules in the target standard structured data to conduct an actual review of the target material text and obtain an actual review result; 3) calling the comparison rules in the target standard structured data to compare the target indicator parameters and obtain a comparison result; 4) integrating the pre-review result, actual review result, and comparison result to obtain and output the review result; The construction process of the preset hierarchical rule base includes: acquiring new energy project rule text and tag dimension data; parsing and encapsulating the new energy project rule text and tag dimension data to obtain standard structured data; and 5) storing the standard structured data hierarchically according to the applicable project type tags to obtain the preset hierarchical rule base.

10. A computing device, characterized in that, include: One or more processors; A storage device for storing one or more programs, which, when executed by one or more processors, cause the one or more processors to implement the method as described in any one of claims 1 to 8.

Images