Power transmission and transformation project design data analysis and evaluation method and system

Abstract

The invention discloses a power transmission and transformation project design data analysis and evaluation method and system, which are applied to the technical field of electric power engineering.The method comprises the steps that multi-source heterogeneous data are obtained from design document data associated with a power transmission and transformation project through recognition, and the multi-source heterogeneous data comprise first-class structured data and second-class unstructured data; performing standardization processing on the first type of structured data to obtain a first type of numerical value feature vector; performing deep analysis on the second type of unstructured data based on a structured semantic analysis rule to obtain a second type of semantic feature vector; performing alignment fusion processing on the first type of numerical value feature vector and the second type of semantic feature vector to obtain a comprehensive feature vector; and inputting the comprehensive feature vector into a pre-constructed design quality evaluation model to obtain a comprehensive evaluation result which is used for adjusting design document data associated with the power transmission and transformation project. According to the method, efficient and objective evaluation of the power transmission and transformation project is realized.

Description

Technical Field

[0001] This invention relates to the field of power engineering technology, and in particular to a method and system for analyzing and evaluating design data in power transmission and transformation projects. Background Technology

[0002] The construction and operation quality of power transmission and transformation projects are the fundamental guarantee for the safe, reliable, and efficient power supply of the power grid. Therefore, conducting a scientific, accurate, and comprehensive evaluation of power transmission and transformation projects is a key link in timely identifying potential hidden dangers, guiding operation and maintenance decisions, optimizing resource allocation, and ensuring the benefits of the project throughout its entire life cycle.

[0003] In existing technologies, during the design of power transmission and transformation projects, reviewers examine a series of text documents related to the project to obtain manual evaluation results. However, as the scale of power transmission and transformation projects becomes increasingly large, on the one hand, the efficiency of manual review is insufficient to meet the evaluation schedule requirements; on the other hand, the review results are influenced by a significant amount of subjective factors, making it difficult to accurately reflect the quality evaluation results of the power transmission and transformation projects. Summary of the Invention

[0004] This invention provides a method and system for analyzing and evaluating design data of power transmission and transformation projects, in order to solve the technical problems of one-sided and inefficient manual evaluation results, and to achieve efficient and objective evaluation of power transmission and transformation projects.

[0005] To address the aforementioned technical problems, this invention provides, in one aspect, a method for analyzing and evaluating design data in power transmission and transformation projects, comprising: Multi-source heterogeneous data is identified from design document data associated with power transmission and transformation projects. The multi-source heterogeneous data includes a first type of structured data and a second type of unstructured data. The first type of structured data is standardized to obtain the first type of numerical feature vector; Determine the structured semantic parsing rules corresponding to the set evaluation criteria, and perform deep parsing on the second type of unstructured data based on the structured semantic parsing rules to obtain the second type of semantic feature vector; The first type of numerical feature vector and the second type of semantic feature vector are aligned and fused to obtain a comprehensive feature vector. The comprehensive feature vector is input into the pre-constructed design quality evaluation model to obtain a comprehensive evaluation result, which is used to adjust the design document data associated with the power transmission and transformation project.

[0006] As one preferred embodiment, the determination of structured semantic parsing rules corresponding to the established evaluation criteria, and the deep parsing of the second type of unstructured data based on the structured semantic parsing rules to obtain the second type of semantic feature vector, including: The structured semantic parsing rules are determined based on predefined design evaluation criteria; Based on the structured semantic parsing rules, multi-granularity semantic parsing and information extraction processing are performed on the second type of unstructured data to obtain structured semantic representation results; The structured semantic representation result is subjected to vector transformation processing to obtain the second type of semantic feature vector.

[0007] As one preferred embodiment, the step of aligning and fusing the first type of numerical feature vector and the second type of semantic feature vector to obtain a comprehensive feature vector includes: Spatial alignment is performed on the first type of numerical feature vector and the second type of semantic feature vector to obtain the first projection feature vector and the second projection feature vector. The first projection feature vector and the second projection feature vector are subjected to feature fusion processing to obtain the comprehensive feature vector.

[0008] As one preferred embodiment, the processing procedure of the design quality evaluation model includes: Based on the multi-head attention mechanism, the dependency relationship between any two feature elements in the comprehensive feature vector is extracted, and a nonlinear transformation is performed on all the dependencies to obtain the encoded feature vector. The encoded feature vector is evaluated and analyzed according to the multi-dimensional quality evaluation criteria to obtain the quality evaluation results for each dimension. The quality evaluation results of all dimensions are correlated and aggregated to obtain the comprehensive evaluation result.

[0009] As one preferred embodiment, the comprehensive evaluation results are used to adjust the design document data associated with the power transmission and transformation project, including: The comprehensive evaluation results are processed to determine defects, thereby obtaining quality defect information in various dimensions. The quality defect information for each dimension is traced and located to obtain an adjustment plan for the specific design item; The design document data associated with the power transmission and transformation project is adjusted based on the aforementioned adjustment scheme.

[0010] Another aspect of the present invention provides a power transmission and transformation engineering design data analysis and evaluation system, comprising: The data acquisition module is used to identify multi-source heterogeneous data from design document data associated with power transmission and transformation projects. The multi-source heterogeneous data includes a first type of structured data and a second type of unstructured data. The standardization processing module is used to standardize the first type of structured data to obtain the first type of numerical feature vector; The semantic parsing module is used to determine the structured semantic parsing rules corresponding to the set evaluation criteria, and to perform deep parsing on the second type of unstructured data based on the structured semantic parsing rules to obtain the second type of semantic feature vector; The fusion processing module is used to align and fuse the first type of numerical feature vector and the second type of semantic feature vector to obtain a comprehensive feature vector. The evaluation result generation module is used to input the comprehensive feature vector into the pre-constructed design quality evaluation model to obtain a comprehensive evaluation result. The comprehensive evaluation result is used to adjust the design document data associated with the power transmission and transformation project.

[0011] As one preferred embodiment, the semantic parsing module includes: A rule-determining unit is used to determine the structured semantic parsing rules based on predefined design evaluation criteria; The semantic parsing unit is used to perform multi-granular semantic parsing and information extraction processing on the second type of unstructured data based on the structured semantic parsing rules, so as to obtain the structured semantic representation result; The vector transformation unit is used to perform vector transformation processing on the structured semantic representation result to obtain the second type of semantic feature vector.

[0012] As one preferred embodiment, the fusion processing module includes: A spatial alignment processing unit is used to perform spatial alignment processing on the first type of numerical feature vector and the second type of semantic feature vector to obtain a first projection feature vector and a second projection feature vector. The feature fusion processing unit is used to perform feature fusion processing on the first projection feature vector and the second projection feature vector to obtain the comprehensive feature vector.

[0013] As one preferred embodiment, the processing procedure of the design quality evaluation model includes: Based on the multi-head attention mechanism, the dependency relationship between any two feature elements in the comprehensive feature vector is extracted, and a nonlinear transformation is performed on all the dependencies to obtain the encoded feature vector. The encoded feature vector is evaluated and analyzed according to the multi-dimensional quality evaluation criteria to obtain the quality evaluation results for each dimension. The quality evaluation results of all dimensions are correlated and aggregated to obtain the comprehensive evaluation result.

[0014] As one preferred embodiment, the evaluation result generation module includes: The defect determination and processing unit is used to perform defect determination processing on the comprehensive evaluation results to obtain quality defect information in various dimensions. The source tracing and location processing unit is used to perform source tracing and location processing on the quality defect information of each dimension to obtain an adjustment scheme for a specific design item. An adjustment unit is used to adjust the design document data associated with the power transmission and transformation project based on the adjustment scheme.

[0015] Compared with the prior art, the beneficial effects of the embodiments of the present invention are at least one of the following: This application achieves a structured and unified expression of multi-source heterogeneous data by standardizing the first type of structured data to generate numerical feature vectors and performing semantic parsing and feature extraction on the second type of unstructured data to generate semantic feature vectors; This application achieves collaborative representation of multi-source data in the feature space by aligning and fusing the numerical feature vectors and semantic feature vectors to obtain a comprehensive feature vector; This application achieves an objective and quantitative evaluation of the design scheme by inputting the comprehensive feature vector into a pre-constructed design quality evaluation model and automatically generating a comprehensive evaluation result, thus solving the problem of inconsistent evaluation standards and difficulty in accurately reflecting the quality evaluation results of power transmission and transformation projects due to reliance on expert subjective experience. Attached Figure Description

[0016] Figure 1 This is a flowchart illustrating a method for analyzing and evaluating power transmission and transformation engineering design data in one embodiment of the present invention. Figure 2 This is a schematic diagram of the structure of a power transmission and transformation engineering design data analysis and evaluation system in one embodiment of the present invention; Figure label: The module includes: 11. Data acquisition module; 12. Standardization processing module; 13. Semantic parsing module; 14. Fusion processing module; and 15. Evaluation result generation module. Detailed Implementation

[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The purpose of providing these embodiments is to make the disclosure of the present invention more thorough and comprehensive. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0018] In the description of this application, the terms "first," "second," "third," etc., are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Therefore, a feature defined with "first," "second," "third," etc., may explicitly or implicitly include one or more of that feature. In the description of this application, unless otherwise stated, "a plurality of" means two or more.

[0019] In the description of this application, it should be noted that, unless otherwise expressly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to fixed connections, detachable connections, or integral connections; they can refer to mechanical connections or electrical connections; they can refer to direct connections or indirect connections through an intermediate medium; and they can refer to the internal communication between two components. The terms "vertical," "horizontal," "left," "right," "upper," "lower," and similar expressions used herein are for illustrative purposes only and do not indicate or imply that the device or component referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as limiting the invention. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items. Those skilled in the art can understand the specific meaning of the above terms in this application based on the specific circumstances.

[0020] In the description of this application, it should be noted that, unless otherwise defined, all technical and scientific terms used in this invention have the same meaning as commonly understood by one of ordinary skill in the art. The terminology used in this specification is for the purpose of describing specific embodiments only and is not intended to limit the invention. Those skilled in the art can understand the specific meaning of the above terms in this application based on the specific circumstances.

[0021] Power transmission and transformation projects form the physical backbone of a new power system, and the quality of their planning, design, and construction fundamentally determines the safety, reliability, economy, and environmental friendliness of the power grid's long-term operation. Therefore, establishing a scientific, accurate, and life-cycle-wide engineering evaluation system is of indispensable strategic significance for achieving early warning of potential risks, guiding precise operation and maintenance decisions, optimizing the allocation of massive assets, and even enhancing the overall resilience and efficiency of the power grid.

[0022] In the critical stage of power transmission and transformation engineering design, which determines subsequent costs and quality, the evaluation of design quality heavily relies on traditional manual review methods. Specifically, reviewers manually examine a series of text documents and charts in various formats and with scattered information, including electrical primary / secondary design specifications, equipment and material lists, floor plans, cost estimates, and calculation sheets. The review process primarily relies on the reviewers' personal professional knowledge and experience, using visual inspection, manual verification, and experiential judgment to attempt to identify omissions, errors, or discrepancies with regulations in the design, ultimately forming a qualitative or semi-quantitative manual evaluation conclusion. However, this evaluation method, dominated by human experience and centered on qualitative judgment, is no longer adequate in terms of efficiency, objectivity, depth, and operability to meet the urgent needs of modern power transmission and transformation engineering's refined and intelligent design and management.

[0023] One embodiment of the present invention provides a method for analyzing and evaluating design data in power transmission and transformation projects. For details, please refer to [link / reference]. Figure 1 , Figure 1 The diagram shown is a flowchart illustrating a data analysis and evaluation method for power transmission and transformation engineering design, according to one embodiment of the present invention. The method includes steps S1 to S5: S1. Identify multi-source heterogeneous data from design document data associated with power transmission and transformation projects. The multi-source heterogeneous data includes a first type of structured data and a second type of unstructured data. S2. Standardize the first type of structured data to obtain the first type of numerical feature vector; S3. Determine the structured semantic parsing rules corresponding to the set evaluation criteria, and perform deep parsing on the second type of unstructured data based on the structured semantic parsing rules to obtain the second type of semantic feature vector; S4. Align and fuse the first type of numerical feature vector and the second type of semantic feature vector to obtain a comprehensive feature vector; S5. Input the comprehensive feature vector into the pre-constructed design quality evaluation model to obtain the comprehensive evaluation result. The comprehensive evaluation result is used to adjust the design document data associated with the power transmission and transformation project.

[0024] Traditional methods relying on manual review and extraction are not only inefficient but also suffer from inconsistent data extraction standards due to differences in individual understanding, making it difficult to support large-scale, standardized, and reproducible intelligent analysis. In this invention, raw and disorganized information from various design documents is processed through a pre-defined, automated technical process, replacing manual operations and ensuring the speed, scale, and consistency of data identification.

[0025] Furthermore, in step S1, during the construction of the intelligent evaluation system for power transmission and transformation projects, automatically identifying and acquiring multi-source heterogeneous data from the design document data associated with the project is the primary and crucial technical step for achieving accurate and efficient design quality evaluation. Specifically, a modular pipeline-style technical architecture is adopted. First, the document parsing and preprocessing module is responsible for connecting to original design documents in different formats (such as PDF, CAD, Word, Excel, etc.), using technologies such as format parsing engines, optical character recognition (OCR), and vector graphics parsing to uniformly deconstruct the document content into basic elements such as text streams, tabular data, images, and metadata. Second, the multimodal data recognition and classification module, based on a pre-built knowledge graph and data tagging system in the field of power transmission and transformation projects, intelligently identifies and classifies the parsed basic elements. For example, it uses natural language processing (NLP) models to identify the technical chapter to which a text paragraph belongs (such as "insulation coordination design instructions"), uses computer vision (CV) algorithms to determine whether an image is a "main wiring diagram" or an "equipment layout diagram," and automatically identifies and extracts well-formatted tabular data into structured data to be processed. This process is highly modular, with each functional module having a clear responsibility, capable of independent optimization and upgrades, and interconnected through standard interfaces to collaboratively complete the transformation from raw documents to standardized data objects. Ultimately, the system outputs clearly labeled and categorized first-class structured data (such as equipment parameter tables and performance index lists) and second-class unstructured data (such as design specification texts, technical demonstration descriptions, and design intent sketches).

[0026] The advantage of this step is that it lays a comprehensive data foundation for design quality evaluation. Through mechanized and automated identification, all potential information sources in a massive amount of design documents can be covered without omission, ensuring the integrity of the evaluation criteria and overcoming the limitations of manual sampling.

[0027] Furthermore, in step S2, the structured data in power transmission and transformation engineering design, such as equipment electrical parameters, material specifications and dimensions, bill of quantities, and technical and economic indicators, although existing in tabular or field form, come from diverse sources, have different dimensions (e.g., voltage unit kV, current unit A, length unit m), and significantly different numerical magnitudes (e.g., cost amount versus equipment quantity), and may contain missing, abnormal, or inconsistent formats. If these raw data are directly input into the model, the differences in their dimensions and numerical distributions will severely interfere with model training and convergence, leading to distorted evaluation results. Therefore, a mechanized preprocessing procedure is necessary to transform this data into a clean, standardized, and mathematically operable unified numerical representation.

[0028] Specifically, firstly, the data cleaning and validation module scans and processes missing values, and identifies and corrects obvious outliers. Next, the core data standardization module is activated, selecting an appropriate algorithm based on the data characteristics. Preferably, for parameters with clear upper and lower bounds, Min-Max normalization is used to linearly map them to the [0,1] interval; for data conforming to a normal distribution, Z-Score standardization is used to make the mean 0 and the standard deviation 1; for data with power-law distribution characteristics, logarithmic transformation may be used. Finally, the feature vector assembly module is responsible for concatenating and combining all the processed structured fields in a predefined order to form the first type of numerical feature vector.

[0029] Relying on the experience of reviewers for subjective interpretation only allows for simple keyword matching of text, which cannot guarantee the comprehensiveness and consistency of the interpretation, nor can it capture the implicit semantics of complex technical logic, compliance arguments, and design flaws. Therefore, further, in step S3, a mechanized semantic parsing framework based on domain knowledge is established to transform the design ideas, technical arguments, and normative references described in human natural language into standardized semantic representations that can be computed and reasoned by machines, thereby providing key input for the comprehensive evaluation that integrates numerical and semantic information.

[0030] Specifically, firstly, structured semantic parsing rules are determined based on predefined design evaluation criteria. This stage forms the foundation for building semantic understanding capabilities. The predefined design evaluation criteria constitute a systematic and modular knowledge system, whose core sources include: mandatory design codes and standards promulgated by the state and industry, a knowledge base of typical design defects and optimization cases accumulated in historical projects, and formally organized domain expert experience criteria. Secondly, based on the aforementioned structured semantic parsing rules, multi-granularity semantic parsing and information extraction processing are performed on the second type of unstructured data to obtain structured semantic representation results. Multi-granularity refers to the parsing process being conducted collaboratively at multiple levels, including vocabulary, sentences, paragraphs, and even document chapters. At the vocabulary level, named entity recognition and key term classification are performed; at the sentence level, dependency parsing is performed to understand the complete logical structure of technical requirements, and the aforementioned rules are used for pattern matching and information extraction; at a higher level, contextual coherence and argument completeness are analyzed. For illustrations and annotations in design drawings, optical character recognition and image understanding technologies are used for initial conversion before incorporating them into the text parsing process. The collection of all such units constitutes the intermediate structured semantic representation. Finally, the structured semantic representation is transformed into a vector vector to obtain the second type of semantic feature vector. Preferably, this transformation process utilizes a pre-trained language model, taking each structured representation unit (or its key field combination) as input and converting it into a high-dimensional semantic vector using a BERT encoder model. Subsequently, a hierarchical pooling aggregation method is used to fuse and compress the vectors of all units, ultimately generating a fixed-dimensional, dense second-type semantic feature vector that comprehensively represents the core semantic information of the entire unstructured data block.

[0031] In the field of power transmission and transformation engineering design, objective numerical parameters (such as equipment specifications and electrical calculation values) and subjective semantic descriptions (such as design specifications and argumentation logic) carry different types of complementary quality information, existing in completely heterogeneous mathematical and semantic spaces. If these are simply spliced ​​together or analyzed independently, the evaluation model will fail to establish the intrinsic connection between numerical anomalies and textual description defects, resulting in a superficial or one-sided evaluation. Therefore, further, in step S4, these two heterogeneous features are deeply aligned and their information is complemented within a unified framework to generate a high-order feature representation that combines objective accuracy with subjective logic, providing a unique and authoritative data basis for making a comprehensive and accurate design quality evaluation decision.

[0032] Specifically, firstly, through two independent linear transformation layers, numerical and semantic feature vectors are projected onto a pre-defined, shared-dimensional common latent space, mathematically resolving the "incomparability" problem caused by their different dimensions and distributions, and outputting a preliminarily aligned projection vector. Secondly, a bidirectional cross-attention mechanism acts as the engine, allowing numerical features to act as "queries" to actively "inquire" and aggregate the most relevant semantic context; then, semantic features act as "queries" to reversely "search" for specific numerical evidence supporting their arguments. This process generates semantically enhanced numerical features and numerically enhanced semantic features, achieving bidirectional information injection and deep alignment. Finally, the two interacting features are concatenated and input into a multilayer perceptron (MLP) consisting of fully connected layers and activation functions, outputting a refined and dense comprehensive feature vector.

[0033] Further, in step S5, the high-dimensional information essence (i.e., the comprehensive feature vector) extracted from the preceding steps, which deeply integrates objective numerical values ​​and subjective semantics, is transformed into quantitative conclusions and action guidelines with clear engineering semantics through a mechanized intelligent reasoning engine. Specifically, the output of the comprehensive evaluation result depends on a design quality evaluation model. Preferably, the design quality evaluation model is a hierarchical deep neural network architecture, including a feature enhancement and understanding module, a multi-dimensional parallel evaluation module, and a comprehensive decision-making module. The feature enhancement and understanding module uses a multi-head attention mechanism to extract the dependency relationship between any two feature elements in the input comprehensive feature vector and performs a nonlinear transformation on all dependencies to obtain the encoded feature vector. The multi-dimensional parallel evaluation module consists of multiple parallel sub-networks, each of which is specifically responsible for a predefined quality dimension, evaluating and analyzing the encoded feature vector to obtain the quality evaluation result for each dimension. The comprehensive decision-making module performs correlation and aggregation processing on the quality evaluation results of all dimensions to obtain the comprehensive evaluation result.

[0034] Preferably, the process of adjusting the design document data related to power transmission and transformation projects based on the comprehensive evaluation results includes: First, parsing the structured comprehensive evaluation results, calling the pre-set defect judgment rule base, automatically identifying dimensions with scores below the threshold, and extracting the feature contribution data built into the model to generate structured quality defect information that clearly points to specific problems. Second, based on the feature index contained in the defect information, tracing back to the original data fragment on which the feature was generated, and obtaining an adjustment scheme for the specific design item. Finally, this adjustment scheme is integrated into the collaborative design platform via API to adjust the design document data.

[0035] Another embodiment of the present invention provides a power transmission and transformation engineering design data analysis and evaluation system. For details, please refer to [link to relevant documentation]. Figure 2 , Figure 2The diagram shown illustrates the structure of a power transmission and transformation engineering design data analysis and evaluation system according to one embodiment of the present invention. The system includes: The data acquisition module is used to identify multi-source heterogeneous data from design document data associated with power transmission and transformation projects. The multi-source heterogeneous data includes a first type of structured data and a second type of unstructured data. The standardization processing module is used to standardize the first type of structured data to obtain the first type of numerical feature vector; The semantic parsing module is used to determine the structured semantic parsing rules corresponding to the set evaluation criteria, and to perform deep parsing on the second type of unstructured data based on the structured semantic parsing rules to obtain the second type of semantic feature vector; The fusion processing module is used to align and fuse the first type of numerical feature vector and the second type of semantic feature vector to obtain a comprehensive feature vector. The evaluation result generation module is used to input the comprehensive feature vector into the pre-constructed design quality evaluation model to obtain a comprehensive evaluation result. The comprehensive evaluation result is used to adjust the design document data associated with the power transmission and transformation project.

[0036] Furthermore, in the above embodiments, the semantic parsing module includes: A rule-determining unit is used to determine the structured semantic parsing rules based on predefined design evaluation criteria; The semantic parsing unit is used to perform multi-granular semantic parsing and information extraction processing on the second type of unstructured data based on the structured semantic parsing rules, so as to obtain the structured semantic representation result; The vector transformation unit is used to perform vector transformation processing on the structured semantic representation result to obtain the second type of semantic feature vector.

[0037] Furthermore, in the above embodiments, the fusion processing module includes: A spatial alignment processing unit is used to perform spatial alignment processing on the first type of numerical feature vector and the second type of semantic feature vector to obtain a first projection feature vector and a second projection feature vector. The feature fusion processing unit is used to perform feature fusion processing on the first projection feature vector and the second projection feature vector to obtain the comprehensive feature vector.

[0038] Furthermore, in the above embodiments, the process of designing the quality evaluation model includes: Based on the multi-head attention mechanism, the dependency relationship between any two feature elements in the comprehensive feature vector is extracted, and a nonlinear transformation is performed on all the dependencies to obtain the encoded feature vector. The encoded feature vector is evaluated and analyzed according to the multi-dimensional quality evaluation criteria to obtain the quality evaluation results for each dimension. The quality evaluation results of all dimensions are correlated and aggregated to obtain the comprehensive evaluation result.

[0039] Furthermore, in the above embodiments, the evaluation result generation module includes: The defect determination and processing unit is used to perform defect determination processing on the comprehensive evaluation results to obtain quality defect information in various dimensions. The source tracing and location processing unit is used to perform source tracing and location processing on the quality defect information of each dimension to obtain an adjustment scheme for a specific design item. An adjustment unit is used to adjust the design document data associated with the power transmission and transformation project based on the adjustment scheme.

[0040] Compared with the prior art, the beneficial effects of the embodiments of the present invention are at least one of the following: (1) This application achieves a structured and unified expression of multi-source heterogeneous data by standardizing the first type of structured data to generate numerical feature vectors and performing semantic parsing and feature extraction on the second type of unstructured data to generate semantic feature vectors. (2) This application achieves the collaborative representation of multi-source data in the feature space by aligning and fusing numerical feature vectors and semantic feature vectors to obtain a comprehensive feature vector; (3) This application automatically generates comprehensive evaluation results by inputting the comprehensive feature vector into the pre-constructed design quality evaluation model, thereby realizing an objective and quantitative evaluation of the design scheme and solving the problem that relying on the subjective experience of experts leads to inconsistent evaluation standards and difficulty in accurately reflecting the quality evaluation results of power transmission and transformation projects.

[0041] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention. Therefore, the scope of protection of this patent should be determined by the appended claims.

Claims

1. A power transmission project design data analysis and evaluation method, characterized in that, The method comprises the following steps: identifying multi-source heterogeneous data from design document data associated with a power transmission and transformation project, wherein the multi-source heterogeneous data comprises first structured data and second unstructured data; standardizing the first structured data to obtain a first numerical feature vector; determining a structured semantic analysis rule corresponding to a set evaluation standard, and performing deep analysis on the second unstructured data based on the structured semantic analysis rule to obtain a second semantic feature vector; aligning and fusing the first numerical feature vector and the second semantic feature vector to obtain a comprehensive feature vector; inputting the comprehensive feature vector into a pre-constructed design quality evaluation model to obtain a comprehensive evaluation result, which is used to adjust the design document data associated with the power transmission and transformation project.

2. The power transmission project design data analysis and evaluation method of claim 1, wherein, The method comprises the following steps: determining the structured semantic analysis rule based on a pre-defined design evaluation basis; performing multi-granularity semantic analysis and information extraction processing on the second unstructured data based on the structured semantic analysis rule to obtain a structured semantic representation result; performing vector conversion processing on the structured semantic representation result to obtain the second semantic feature vector.

3. The method for analyzing and evaluating design data of power transmission and transformation projects as described in claim 1, characterized in that, The method comprises the following steps: performing spatial alignment processing on the first numerical feature vector and the second semantic feature vector to obtain a first projection feature vector and a second projection feature vector; performing feature fusion processing on the first projection feature vector and the second projection feature vector to obtain the comprehensive feature vector.

4. The method for analyzing and evaluating design data of power transmission and transformation projects as described in claim 1, characterized in that, The processing process of the design quality evaluation model comprises the following steps: extracting dependency relationships between any two feature elements in the comprehensive feature vector based on a multi-head attention mechanism, and performing nonlinear transformation on all the dependency relationships to obtain an encoded feature vector; performing evaluation analysis on the encoded feature vector according to a multi-dimensional quality evaluation standard to obtain a quality evaluation result of each dimension; performing correlation aggregation processing on the quality evaluation results of all dimensions to obtain the comprehensive evaluation result.

5. The method for analyzing and evaluating design data of power transmission and transformation projects as described in claim 1, characterized in that, The comprehensive evaluation result is used to adjust the design document data associated with the power transmission and transformation project, which comprises the following steps: performing defect judgment processing on the comprehensive evaluation result to obtain quality defect information of each dimension; performing traceability positioning processing on the quality defect information of each dimension to obtain an adjustment scheme for a specific design item; adjusting the design document data associated with the power transmission and transformation project based on the adjustment scheme.

6. A power transmission project design data analysis and evaluation system, characterized by, The method comprises the following steps: a data acquisition module is configured to identify multi-source heterogeneous data from design document data associated with a power transmission and transformation project, wherein the multi-source heterogeneous data comprises first structured data and second unstructured data; The standardized processing module is configured to perform standardized processing on the first structured data to obtain a first numerical feature vector; The semantic analysis module is configured to determine a structured semantic analysis rule corresponding to a set evaluation standard, perform deep analysis on the second unstructured data based on the structured semantic analysis rule, and obtain a second semantic feature vector; The fusion processing module is configured to perform alignment and fusion processing on the first numerical feature vector and the second semantic feature vector to obtain a comprehensive feature vector; The evaluation result generation module is configured to input the comprehensive feature vector into a pre-constructed design quality evaluation model to obtain a comprehensive evaluation result, which is used to adjust the design document data associated with the power transmission and transformation project.

7. The power transmission project design data analysis and evaluation system of claim 6, wherein, The semantic analysis module includes: The determination analysis rule unit is configured to determine the structured semantic analysis rule based on a pre-defined design evaluation basis; The semantic analysis unit is configured to perform multi-granularity semantic analysis and information extraction processing on the second unstructured data based on the structured semantic analysis rule to obtain a structured semantic representation result; The vector conversion unit is configured to perform vector conversion processing on the structured semantic representation result to obtain the second semantic feature vector.

8. The power transmission project design data analysis and evaluation system of claim 6, wherein, The fusion processing module includes: The spatial alignment processing unit is configured to perform spatial alignment processing on the first numerical feature vector and the second semantic feature vector to obtain a first projection feature vector and a second projection feature vector; The feature fusion processing unit is configured to perform feature fusion processing on the first projection feature vector and the second projection feature vector to obtain the comprehensive feature vector.

9. The power transmission project design data analysis and evaluation system of claim 6, wherein, The processing process of the design quality evaluation model includes: Based on the multi-head attention mechanism, the dependency relationship between any two feature elements in the comprehensive feature vector is extracted, and all the dependency relationships are nonlinearly transformed to obtain an encoded feature vector; According to the multi-dimensional quality evaluation standard, the encoded feature vector is evaluated and analyzed to obtain a quality evaluation result of each dimension; The quality evaluation results of all dimensions are associated and aggregated to obtain the comprehensive evaluation result.

10. The power transmission project design data analysis and evaluation system of claim 6, wherein, The evaluation result generation module includes: The defect judgment processing unit is configured to perform defect judgment processing on the comprehensive evaluation result to obtain quality defect information of each dimension; The trace positioning processing unit is configured to perform trace positioning processing on the quality defect information of each dimension to obtain an adjustment scheme for a specific design item; The adjustment unit is configured to adjust the design document data associated with the power transmission and transformation project based on the adjustment scheme.

Images