Multi-dimensional user demand analysis system and method in engineering construction stage

Abstract

The invention relates to an engineering construction stage user multi-dimensional demand analysis system and method, and the system comprises a multi-modal semantic model construction module which integrates a multi-source knowledge base, and extracts triple description statements from the provisions of the multi-source knowledge base to construct an engineering object ontology knowledge graph; the demand structured guidance and intelligent analysis engine module dynamically identifies an engineering stage and a user role, guides a user to perform text demand input, and performs semantic completion and standardization processing; the demand conflict detection and coordination correction module is used for detecting conflicts in user demands and generating a conflict influence analysis report and a coordination processing scheme; the intelligent drawing and dynamic optimization module is used for automatically drawing according to the triple obtained through demand analysis; and the demand analysis visualization and flow closed-loop module is used for generating an executable parameterized demand document and an association chart, and completing demand-design-verification engineering closed-loop management, so that intelligent analysis and standardization of user multi-dimensional demands are realized, and the system has the advantages of high efficiency, accuracy and intelligence.

Description

Technical Field

[0001] This invention belongs to the fields of requirements engineering and artificial intelligence technology, and specifically relates to a system and method for analyzing multi-dimensional user requirements during the engineering construction phase. Background Technology

[0002] Information technology and digitalization are essential components of engineering construction. User needs are the core basis for system design, implementation, development, and acceptance in information technology and digitalization. As the complexity, quantity, and breadth of information systems increase, user needs have become more multi-dimensional, dynamic, and strongly coupled.

[0003] Manual interviews and raw data recording: Implementation personnel conduct manual interviews with users to record raw requirements based on user feedback. They also investigate the original records related to the requirements, break them down and understand them, and then reconfirm them with the users. Finally, implementation personnel manually design and verify the charts. After recording the requirements, implementation personnel need to repeatedly confirm them with users. During this confirmation process, they often encounter ambiguity in requirement expression, imprecise engineering language, conflicts and coordination between personalized requirements and standard project management frameworks, and the varying depth and perspective of requirements at different levels and stages, necessitating further transformation. Furthermore, traditional methods cannot meet timeliness requirements and may pose a risk to project progress.

[0004] Therefore, this application is hereby submitted.

[0005] Furthermore, on the one hand, there are differences in understanding among those skilled in the art; on the other hand, the applicant studied a large number of documents and patents when making this invention, but due to space limitations, not all details and contents were listed in detail. However, this does not mean that the present invention does not possess the features of these prior art. On the contrary, the present invention already possesses all the features of the prior art, and the applicant reserves the right to add relevant prior art to the background art. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this invention provides a system and method for analyzing multi-dimensional user requirements during the engineering construction phase. It enables intelligent analysis and standardization of multi-dimensional user requirements, generates executable parameterized requirement documents and related diagrams, and ultimately forms a closed-loop management process for requirement engineering, including requirement-design-verification, while also enabling timely adjustments to engineering requirements.

[0007] This invention discloses a system for analyzing multi-dimensional user requirements during the engineering construction phase, comprising: The multimodal semantic model construction module is used to build a multimodal semantic model for the engineering construction phase. The multimodal semantic model integrates a multi-source knowledge base and extracts entity, attribute, and relation triple description statements from the articles of the multi-source knowledge base to construct an engineering object ontology knowledge graph. The requirement structuring guidance and intelligent parsing engine module is used to dynamically identify the engineering stage and user role, guide users to input text requirements, and perform semantic completion and standardization processing on the user input based on the engineering object ontology knowledge graph. The Requirement Conflict Detection and Coordination Correction Module is used to detect conflicts in user requirements based on a multi-dimensional rule base built from a multi-source knowledge base, and generate conflict impact analysis reports and coordination solutions. The intelligent drawing and dynamic optimization module is used to automatically generate ER diagrams of engineering objects, interactive prototype diagrams, and workflow diagrams based on the entity, attribute, and relationship triples obtained from the requirements analysis. The requirement analysis visualization and process closure module is used to generate executable parameterized requirement documents and related diagrams to complete the closed-loop management of the project from requirement to design to verification.

[0008] In some specific implementations, constructing a multimodal semantic model for the engineering construction phase includes: Feature extraction is performed on charts in a multi-source knowledge base to generate a topology diagram of engineering object structure; Based on dependency parsing, triplet description statements are automatically extracted from knowledge base entries to construct an engineering object ontology knowledge graph. And / or, set up an incremental learning module to trigger online fine-tuning of the model when new engineering specifications or rule changes are added, updating only the weights of the affected ontology nodes.

[0009] In some specific implementations, the demand structuring guidance and intelligent parsing engine module performs the following operations: The system guides users to input their textual requirements according to the logical chain of "project object, attribute constraint, functional hierarchy, and approval process". Based on the user-uploaded requirement sketches, the sketch type is identified and parsed into standardized parameters, and historical cases and related specification items are automatically matched. Based on the engineering object ontology knowledge graph, the system uses reasoning methods to parse user requirements, extract core engineering objects and attribute constraint tables, and automatically establish object relationships to generate a triple graph.

[0010] In some specific implementations, the demand structuring guidance and intelligent parsing engine module is also configured as follows: Define input restrictions for object attributes based on user permission levels, including input type, regular expressions, and numerical range restrictions; Attribute and function classification is performed by analyzing behavioral verbs in user needs; Based on the knowledge graph of the engineering object ontology, semantic matching is performed to push standardized expression suggestions.

[0011] In some specific implementations, the demand conflict detection and coordination correction module performs the following operations: A multi-dimensional rule base is built based on the knowledge base to quantify the types and levels of rules; Detect numerical logic conflicts, standard compliance conflicts, and attribute description conflicts in user requirements; Automatically merge redundant data and descriptions to generate conflict impact analysis reports and coordination solutions; Users can manually correct conflicting parameters, and the system will synchronously update the associated fields.

[0012] In some specific implementations, the intelligent drawing and dynamic optimization module performs the following operations: Based on the entity relationship description obtained from the requirements analysis, extract the entity, attribute, and relationship triplet, and generate the ER diagram of the engineering object by combining it with the knowledge base. Based on the functional requirements and hierarchies obtained from the requirements analysis, and combined with the system's predefined fields and components, an interactive prototype diagram is generated. Based on the key verbs of the workflow, the responsible personnel, and the requirements for work nodes obtained from the requirements analysis, a workflow diagram is generated.

[0013] In some specific implementations, the intelligent drawing and dynamic optimization module is also configured as follows: Users can interactively modify the generated drawings, including moving nodes, modifying node connection lines, and modifying node properties; After the user completes the modifications, the drawing layout is automatically optimized and the logical integrity is verified based on the engineering object ontology knowledge graph, while the requirement description is updated in reverse. Users can supplement their requirements descriptions using natural language, and the generated drawings can be adaptively adjusted based on the engineering object ontology knowledge graph and multi-dimensional rule base.

[0014] In some specific implementations, the requirements analysis visualization and process closure module is configured to perform the following steps: Generate a requirements analysis report, clearly indicating the steps and reference rules for requirements analysis; Record the history of user requirement changes and mark the scope of related impacts; Generate executable parameterized requirements documents and related diagrams, and upload them to the corresponding digital system to complete the closed loop of the requirements-design-verification process.

[0015] In some specific implementations, the multimodal semantic model building module includes: The machine vision processing unit is used to extract features from the knowledge base graph and generate a structural topology graph. Triple extraction unit is used to construct an engineering object ontology knowledge graph based on dependency parsing; Incremental learning units are used to trigger online fine-tuning of the model when specifications change.

[0016] This invention also discloses a method for analyzing multi-dimensional user requirements during the engineering construction phase, comprising the following steps: A multimodal semantic model for the engineering construction phase is constructed. The multimodal semantic model integrates a multi-source knowledge base and extracts entity, attribute, and relation triplet description statements from the multi-source knowledge base text based on dependency parsing to construct an engineering object ontology knowledge graph. Through demand structuring guidance and intelligent parsing engine, the system dynamically identifies engineering stages and user roles, guides users to input textual requirements, and performs semantic completion and standardization processing on user input based on the engineering object ontology knowledge graph. Based on a multi-dimensional rule base built from a multi-source knowledge base, conflicts in user needs are detected, and conflict impact analysis reports and coordination and handling solutions are generated. Based on the entity, attribute, and relationship triples obtained from the requirements analysis, the system automatically generates an ER diagram of the engineering object, an interactive prototype diagram, and a workflow diagram. Generate executable parameterized requirements documents and related diagrams to complete the closed-loop management of the requirements-design-verification project.

[0017] Beneficial Effects: This solution's system and methodology, by constructing a multimodal semantic model for the engineering construction phase and combining natural language parsing technology with a dynamic rule engine, achieves intelligent parsing, conflict detection, and standardized mapping of users' multidimensional requirements. It generates executable parameterized requirement documents and related diagrams, ultimately forming a closed-loop management process for requirement engineering—from requirement to design to verification. This is particularly suitable for requirement engineering management in EPC general contracting projects in fields such as construction, transportation, and energy. Intelligent parsing and standardized processing resolve issues of ambiguous requirement expressions and insufficiently rigorous engineering language, offering advantages in efficiency, accuracy, and intelligence. Attached Figure Description

[0018] Appendix Figure 1 This is a block diagram illustrating the principle of the method described in this invention; Appendix Figure 2 This is a schematic diagram of the analysis process of a specific embodiment of the present invention. Detailed Implementation

[0019] 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. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0020] Information technology and digitalization are essential components of engineering construction. User needs are the core basis for system design, implementation, development, and acceptance in information technology and digitalization. As the complexity, quantity, and breadth of information systems increase, user needs have become more multi-dimensional, dynamic, and strongly coupled.

[0021] Manual interviews and raw data recording: Implementation personnel conduct manual interviews with users to record raw requirements based on user feedback. They also investigate the original records related to the requirements, break them down and understand them, and then reconfirm them with the users. Finally, implementation personnel manually design and verify the charts. After recording the requirements, implementation personnel need to repeatedly confirm them with users. During this confirmation process, they often encounter ambiguity in requirement expression, imprecise engineering language, conflicts and coordination between personalized requirements and standard project management frameworks, and the varying depth and perspective of requirements at different levels and stages, necessitating further transformation. Furthermore, traditional methods cannot meet timeliness requirements and may pose a risk to project progress.

[0022] Based on this, a multi-dimensional user requirement analysis system for the engineering construction phase is proposed, including: The multimodal semantic model construction module is used to build a multimodal semantic model for the engineering construction phase. This multimodal semantic model integrates a multi-source knowledge base, extracting entity, attribute, and relation triplet descriptions from the multi-source knowledge base to construct an engineering object ontology knowledge graph. The multi-source knowledge base includes at least an engineering specification library, a historical case library, and a commonly used engineering terminology library, providing rich engineering specifications, cases, and engineering terminology.

[0023] The multimodal semantic model for the construction phase of the project includes: Feature extraction is performed on the charts in the multi-source knowledge base to generate a structural topology diagram of engineering objects. Generally, machine vision technology is used to extract features from the charts in the multi-source knowledge base. A semantic hierarchical framework is adopted to establish an engineering object library. A structural topology diagram is generated based on the engineering objects, maintaining consistency between the text and the diagram.

[0024] Based on dependency parsing, triple description statements are automatically extracted from knowledge base entries to construct an engineering object ontology knowledge graph. Triples are the three essential elements of the knowledge graph, including entities, attributes, and relations. By constructing this engineering object ontology knowledge graph and associating various specification parameters of each engineering object, dynamic modeling is performed to form a multimodal semantic model. This enables structured storage and intelligent retrieval of engineering knowledge, supporting semantic-level requirement understanding and analysis.

[0025] In some specific implementations, an incremental learning module can be set up to trigger online fine-tuning of the model when new engineering specifications or rule changes are added, updating only the weights of the affected ontology nodes. The weights range from 0.1 to 1.0 and can be adjusted accordingly as needed. The incremental learning module includes a self-learning model, which is used to train itself based on training data, generated data, etc., thereby further improving the accuracy of the parsing.

[0026] The requirement structuring guidance and intelligent parsing engine module dynamically identifies project stages and user roles, guides users to input textual requirements, and performs semantic completion and standardization processing on the user input based on the project object ontology knowledge graph. Employing a role-based dynamic guidance mechanism, it provides targeted requirement input templates and guidance processes according to different project stages and user identities, reducing the difficulty of user input and improving the completeness of requirement input.

[0027] In some specific implementations, the demand structuring guidance and intelligent parsing engine module performs the following operations: The system guides users to input their textual requirements according to the logical chain of "project object, attribute constraint, functional hierarchy, and approval process". Based on the user-uploaded requirement sketches, the sketch type is identified and parsed into standardized parameters, automatically matching historical cases and related specification items; among them, requirement sketches include: hand-drawn flowcharts, hand-drawn form styles, hand-drawn node details, etc.

[0028] Based on user-uploaded requirement sketches, the system identifies the sketch type and constructs the requirement sketch meaning according to hand-drawn flowcharts, hand-drawn form styles, and hand-drawn node details. It intelligently parses the sketches into standardized parameters and automatically matches historical cases and related specification items. Through sketch recognition and parsing, it supports multimodal requirement input and achieves automatic structured storage of requirements.

[0029] Based on the engineering object ontology knowledge graph, the system uses reasoning methods to parse user requirements, extract core engineering objects and attribute constraint tables, and automatically establish object relationships to generate a triple graph.

[0030] The system analyzes user requirements using logical reasoning, fuzzy reasoning, and case-based reasoning methods, extracts core engineering objects and attribute constraint tables, including numerical restrictions and non-numerical descriptions, automatically establishes object relationships to generate triplet graphs, and stores them in the database.

[0031] In some specific implementations, the demand structuring guidance and intelligent parsing engine module is further configured as follows: Define input restrictions for object attributes based on user permission levels, including input type, regular expressions, and numerical range restrictions; Attribute and function classification is performed by analyzing behavioral verbs in user needs; Based on the knowledge graph of the engineering object ontology, semantic matching is performed to push standardized expression suggestions.

[0032] By leveraging the semantic association features of knowledge graphs, real-time semantic analysis and completion of user input are performed, mapping non-standard expressions to standard engineering terms, improving the standardization and consistency of requirement expressions, reducing misunderstandings caused by inconsistent terminology, and supporting access-controlled requirement input management.

[0033] The requirement conflict detection and coordination module is used to detect conflicts in user requirements based on a multi-dimensional rule base constructed from a multi-source knowledge base, and generate a conflict impact analysis report and coordination and handling plan. In some specific implementations, the requirement conflict detection and coordination module performs the following operations: A multi-dimensional rule base is built based on the knowledge base to quantify the types and levels of rules; Detect numerical logic conflicts, standard compliance conflicts, and attribute description conflicts in user requirements; Automatically merge redundant data and descriptions to generate conflict impact analysis reports and coordination solutions; Users can manually correct conflicting parameters, and the system will synchronously update the associated fields.

[0034] The construction of a multi-dimensional rule base based on a knowledge base includes: Construct numerical logic rules to detect the numerical range, dependencies, and arithmetic logic conflicts among the attributes of engineering objects; Establish standard compliance rules to verify the conformity of user requirements with mandatory provisions in the engineering specification library; Construct attribute description rules to detect terminology consistency, appropriateness of action verb classification, and attribute mutual exclusivity; The rules in the rule base are quantified into categories and levels, and conflicts in user requirements are detected based on this.

[0035] Establish a rule system encompassing three dimensions: numerical logic, standards compliance, and attribute description, with different weights and priorities assigned to each rule type. Enable comprehensive, multi-layered requirement conflict detection; provide quantitative assessment of conflict severity; and support rule customization and extension.

[0036] The intelligent drawing and dynamic optimization module is used to automatically generate ER diagrams of engineering objects, interactive prototype diagrams, and workflow diagrams based on the entity, attribute, and relationship triples obtained from the requirements analysis. The intelligent drawing and dynamic optimization module performs the following operations: Based on the entity relationship description obtained from the requirements analysis, extract the entity, attribute, and relationship triplet, and combine it with the knowledge base to generate an ER diagram of engineering objects, which clearly shows the relationship between objects. Based on the functional requirements and hierarchies obtained from the requirements analysis, and combined with the system's predefined fields and components, an interactive prototype diagram is generated. Based on the key verbs of the workflow, the responsible personnel, and the requirements for work nodes obtained from the requirements analysis, a workflow diagram is generated.

[0037] In some specific implementations, the intelligent drawing and dynamic optimization module is further configured as follows: Users can interactively modify the generated drawings, including moving nodes, modifying node connection lines, and modifying node properties; After the user completes the modifications, the drawing layout is automatically optimized and the logical integrity is verified based on the engineering object ontology knowledge graph, while the requirement description is updated in reverse. Users can supplement their requirements descriptions using natural language, and the generated drawings can be adaptively adjusted based on the engineering object ontology knowledge graph and multi-dimensional rule base.

[0038] Based on the parsed triplet data, the application of a graph layout algorithm automatically generates engineering diagrams, supporting interactive modification and real-time optimization. It automatically generates professional-grade engineering drawings, significantly improving drafting efficiency; supports bidirectional linkage updates between requirements and drawings; and maintains real-time consistency between requirements and design.

[0039] The requirement analysis visualization and process closure module is used to generate executable parameterized requirement documents and related diagrams to complete the closed-loop management of the project from requirement to design to verification.

[0040] The requirements analysis visualization and process closure module is configured to perform the following steps: Generate a requirements analysis report, clearly indicating the steps and reference rules for requirements analysis; Record the history of user requirement changes and mark the scope of related impacts; Generate executable parameterized requirements documents and related diagrams, and upload them to the corresponding digital system to complete the closed loop of the requirements-design-verification process.

[0041] Generate a requirements analysis report, clearly marking the steps and reference rules for requirements analysis, recording the history of user requirement changes, marking the scope of related impacts, and exporting files for subsequent communication or optimization in other systems.

[0042] The multimodal semantic model construction module includes: The machine vision processing unit is used to extract features from the knowledge base graph and generate a structural topology graph. Triple extraction unit is used to construct an engineering object ontology knowledge graph based on dependency parsing; Incremental learning units are used to trigger online fine-tuning of the model when specifications change.

[0043] On the other hand, as attached Figure 1 As shown, a method for analyzing multi-dimensional user requirements during the engineering construction phase is provided, including the following steps: A multimodal semantic model for the engineering construction phase is constructed. The multimodal semantic model integrates a multi-source knowledge base and extracts entity, attribute, and relation triplet description statements from the multi-source knowledge base text based on dependency parsing to construct an engineering object ontology knowledge graph. Through demand structuring guidance and intelligent parsing engine, the system dynamically identifies engineering stages and user roles, guides users to input textual requirements, and performs semantic completion and standardization processing on the user input based on the engineering object ontology knowledge graph. Based on a multi-dimensional rule base built from a multi-source knowledge base, conflicts in user needs are detected, and conflict impact analysis reports and coordination and handling solutions are generated. Based on the entity, attribute, and relationship triples obtained from the requirements analysis, the system automatically generates an ER diagram of the engineering object, an interactive prototype diagram, and a workflow diagram. Generate executable parameterized requirements documents and related diagrams to complete the closed-loop management of the requirements-design-verification project.

[0044] The above-mentioned method for analyzing user multidimensional requirements during the engineering construction phase is implemented through the following modules in the system: multimodal semantic model construction module, requirement structure guidance and intelligent analysis engine module, requirement conflict detection and coordination correction module, intelligent drawing and dynamic optimization module, and requirement analysis visualization and process closure module.

[0045] By integrating multimodal data with text, charts, and sketches, a comprehensive knowledge graph is constructed. Input guidance is dynamically adjusted based on stage and role, reducing user burden. Conflict detection: Based on a multi-dimensional rule base, conflicting requirements are identified in advance. Interactive drawing: When users modify drawings, the system automatically updates requirements, achieving two-way interaction. Closed-loop management: From requirements to design and verification, the entire process is traceable.

[0046] A specific embodiment is provided, as shown in the appendix. Figure 2 The diagram shown is a schematic representation of the analysis process in a specific embodiment.

[0047] This system involves multiple modules, including multimodal semantic model construction, requirement parsing engine, conflict detection, intelligent drawing, and visual closed-loop management. The specific process is as follows: Using an engineering specification library, a common engineering terminology library, and a historical case library as input data, a multi-source knowledge base is formed. Machine vision technology is used to extract features from the graphs in the multi-source knowledge base to generate an engineering object structure topology graph. Based on dependency parsing, entity, attribute, and relation triple description statements are extracted from the articles in the multi-source knowledge base. The resulting topology graph and triple descriptions are used to construct an engineering object ontology knowledge graph, which is then output. An incremental learning unit is also set up to trigger online fine-tuning of the model when rules change or node weights are updated. The model is then output after fine-tuning.

[0048] We will demonstrate how the system works by focusing on a specific engineering construction scenario (e.g., the design and construction of an industrial plant).

[0049] Project Background: An expansion project for a large chemical plant, including the construction of a new reactor building, tank area, control room, etc.

[0050] Users: The project involves multiple roles, including the owner, design institute, construction company, and supervisor.

[0051] Requirements: User requirements include process requirements, safety requirements, operation and maintenance requirements, etc., which are scattered across multiple documents and specifications.

[0052] System architecture implementation: Multimodal semantic model construction module; Data sources: Integrates national building codes, chemical industry standards, enterprise standards, historical project cases, design drawings, etc.

[0053] Knowledge graph construction: Use machine vision processing units to extract features from diagrams in the specifications (such as fire protection system diagrams) to generate structural topology diagrams.

[0054] The triplet extraction unit extracts triples from the specification clauses. For example, it extracts triples from the clause "The reactor plant shall have two safety exits": Entity: Reactor plant; Attribute: Number of safety exits; Relationship: Should be set; Value: 2. The constructed knowledge graph includes engineering objects (such as reactors, storage tanks, and pipelines) and their attributes and relationships.

[0055] Incremental learning: When a new specification is released, the incremental learning unit only updates the affected parts. For example, if the new specification requires that "the width of the safety exit shall not be less than 1.2 meters", then the attributes of the "safety exit" node will be updated.

[0056] Demand Structured Guidance and Intelligent Analysis Engine Module: Dynamically Identifies Engineering Stages and User Roles: The system identifies that the current stage is "Preliminary Design" and the user role is "Process Engineer".

[0057] Guided Input: Following the logical chain of "project object, attribute constraints, functional hierarchy, and approval process," guide users to input their requirements. For example: Project object: reactor; Attribute constraints: operating temperature 300℃, pressure 1.5MPa; Functional hierarchy: core equipment, requiring redundant configuration; Approval process: requires safety review.

[0058] Semantic completion and standardization: When a user inputs "the reactor should be resistant to high temperatures", the system completes the input using a knowledge graph to "the reactor material should be resistant to high temperatures, with an operating temperature range of 0~500℃", and recommends a standardized expression.

[0059] Sketch Analysis: The user uploads a reactor layout sketch, the system identifies the reactor, pipes, and support structures, and parses them into standardized parameters, matching them with historical cases and specification entries.

[0060] Requirement conflict detection and coordination correction module: Rule base: Built on a knowledge base, including rules for numerical logic (such as the relationship between pressure vessel wall thickness and pressure), standard compliance (such as fireproof distance), and attribute description (such as material and temperature compatibility).

[0061] Conflict detection: When the user inputs "reactor material is carbon steel" and "operating temperature 300℃", the system detects that the carbon steel is not strong enough at 300℃, triggering a conflict.

[0062] Report generated: The system generates a conflict impact analysis report, indicating that carbon steel may experience creep at 300℃, and recommends the use of stainless steel.

[0063] Coordination and processing: Users can choose to accept the suggestion, and the system will automatically update the requirements and synchronize the update of related fields (such as cost estimation).

[0064] The intelligent plotting and dynamic optimization module generates charts: ER diagram: An equipment layout ER diagram is generated based on entities such as reactors, pipes, valves, and their relationships.

[0065] Interactive Prototype Diagram: Generates a simplified P&ID diagram of the reactor system based on functional requirements, which users can interactively modify.

[0066] Workflow diagram: Generate a security review workflow diagram based on the approval process requirements.

[0067] Interactive modification: The user drags the reactor position, and the system automatically adjusts the pipe connections and verifies the logic (such as whether the pipe length exceeds the specification).

[0068] Natural language adjustment: When a user says "add a backup pump", the system adds a backup pump in the diagram and updates the requirements document.

[0069] Requirements analysis visualization and process closed-loop module: Generate a requirements analysis report: Record the requirements analysis steps in detail and cite relevant specification clauses.

[0070] Change history: Records changes made by users from initial requirements to final requirements, and marks the scope of impact of each change (such as cost and time).

[0071] Generate an executable parameterized requirements document: Output a structured requirements document, including equipment parameters, layout requirements, process descriptions, etc., and upload it to a project management system (such as Primavera or MS Project) to complete the loop.

[0072] Workflow example: Step 1: The user (process engineer) logs into the system, selects the project stage "Preliminary Design", and the system displays the requirements input interface.

[0073] Step 2: The system guides the user to input reactor requirements: Project object: reactor.

[0074] Attribute constraints: User input "Operating temperature 300℃, pressure 1.5MPa, material carbon steel".

[0075] Functional classification: Select "Core equipment".

[0076] Approval process: Select "Security review required".

[0077] Step 3: The system performs a conflict detection and finds that the carbon steel material conflicts with the 300℃ temperature, prompting the user and suggesting the use of stainless steel. The user accepts the suggestion.

[0078] Step 4: The system generates the ER diagram and interactive prototype diagram of the reactor system according to the requirements. The user can adjust the reactor position interactively, and the system will automatically update the pipeline connection.

[0079] Step 5: The system generates a security review workflow diagram and links it to the project plan.

[0080] Step 6: The system outputs a parameterized requirements document, including all parameters, drawings, and processes of the reactor, and uploads it to the project management system.

[0081] This approach can improve the efficiency of requirement coordination during the engineering construction process, identify requirement conflicts early, reduce the cost of later changes, and improve the accuracy and consistency of requirement documents.

[0082] It should be noted that the specific embodiments described above are exemplary. Those skilled in the art can devise various solutions inspired by the disclosure of this invention, and these solutions all fall within the scope of this invention and its protection. Those skilled in the art should understand that this specification and its accompanying drawings are illustrative and do not constitute a limitation on the claims. The scope of protection of this invention is defined by the claims and their equivalents. This specification contains multiple inventive concepts; phrases such as "preferred" or "according to a preferred embodiment" indicate that the corresponding paragraph discloses an independent concept. The applicant reserves the right to file divisional applications based on each inventive concept. Throughout the text, the feature introduced by "preferred" is only an optional mode and should not be construed as mandatory. Therefore, the applicant reserves the right to abandon or delete relevant preferred features at any time.

Claims

1. A system for analyzing multi-dimensional user needs during the engineering construction phase, characterized in that, include: A multimodal semantic model construction module is used to construct a multimodal semantic model for the engineering construction phase. The multimodal semantic model integrates a multi-source knowledge base and extracts entity, attribute, and relation triple description statements from the articles of the multi-source knowledge base to construct an engineering object ontology knowledge graph. The demand structuring guidance and intelligent parsing engine module is used to dynamically identify the engineering stage and user role, guide the user to input text requirements, and perform semantic completion and standardization processing on the user input based on the engineering object ontology knowledge graph. The Requirement Conflict Detection and Coordination Correction Module is used to detect conflicts in user requirements based on a multi-dimensional rule base built from a multi-source knowledge base, and generate conflict impact analysis reports and coordination solutions. The intelligent drawing and dynamic optimization module is used to automatically generate ER diagrams of engineering objects, interactive prototype diagrams, and workflow diagrams based on the entity, attribute, and relationship triples obtained from the requirements analysis. The requirement analysis visualization and process closure module is used to generate executable parameterized requirement documents and related diagrams to complete the closed-loop management of the project from requirement to design to verification.

2. The system according to claim 1, characterized in that, The multimodal semantic model for the construction phase of the project includes: Feature extraction is performed on charts in a multi-source knowledge base to generate a topology diagram of engineering object structure; Based on dependency parsing, triplet description statements are automatically extracted from knowledge base entries to construct an engineering object ontology knowledge graph. And / or, set up an incremental learning module to trigger online fine-tuning of the model when new engineering specifications or rule changes are added, updating only the weights of the affected ontology nodes.

3. The system according to claim 1 or 2, characterized in that, The demand structuring guidance and intelligent parsing engine module performs the following operations: The system guides users to input their textual requirements according to the logical chain of "project object, attribute constraint, functional hierarchy, and approval process". Based on the user-uploaded requirement sketches, the sketch type is identified and parsed into standardized parameters, and historical cases and related specification items are automatically matched. Based on the engineering object ontology knowledge graph, the system uses reasoning methods to parse user requirements, extract core engineering objects and attribute constraint tables, and automatically establish object relationships to generate a triple graph.

4. The system according to any one of claims 1 to 3, characterized in that, The demand structuring guidance and intelligent parsing engine module is also configured as follows: Define input restrictions for object attributes based on user permission levels, including input type, regular expressions, and numerical range restrictions; Attribute and function classification is performed by analyzing behavioral verbs in user needs; Based on the knowledge graph of the engineering object ontology, semantic matching is performed to push standardized expression suggestions.

5. The system according to any one of claims 1 to 4, characterized in that, The requirement conflict detection and coordination correction module performs the following operations: A multi-dimensional rule base is built based on the knowledge base to quantify the types and levels of rules; Detect numerical logic conflicts, standard compliance conflicts, and attribute description conflicts in user requirements; Automatically merge redundant data and descriptions to generate conflict impact analysis reports and coordination solutions; Users can manually correct conflicting parameters, and the system will synchronously update the associated fields.

6. The system according to any one of claims 1 to 5, characterized in that, The intelligent drawing and dynamic optimization module performs the following operations: Based on the entity relationship description obtained from the requirements analysis, extract the entity, attribute, and relationship triplet, and generate the ER diagram of the engineering object by combining it with the knowledge base. Based on the functional requirements and hierarchies obtained from the requirements analysis, and combined with the system's predefined fields and components, an interactive prototype diagram is generated. Based on the key verbs of the workflow, the responsible personnel, and the requirements for work nodes obtained from the requirements analysis, a workflow diagram is generated.

7. The system according to any one of claims 1 to 6, characterized in that, The intelligent drawing and dynamic optimization module is also configured to: Users can interactively modify the generated drawings, including moving nodes, modifying node connection lines, and modifying node properties; After the user completes the modifications, the drawing layout is automatically optimized and the logical integrity is verified based on the engineering object ontology knowledge graph, while the requirement description is updated in reverse. Users can supplement their requirements descriptions using natural language, and the generated drawings can be adaptively adjusted based on the engineering object ontology knowledge graph and multi-dimensional rule base.

8. The system according to any one of claims 1 to 7, characterized in that, The requirements analysis visualization and process closure module is configured to perform the following steps: Generate a requirements analysis report, clearly indicating the steps and reference rules for requirements analysis; Record the history of user requirement changes and mark the scope of related impacts; Generate executable parameterized requirements documents and related diagrams, and upload them to the corresponding digital system to complete the closed loop of the requirements-design-verification process.

9. The system according to any one of claims 1 to 8, characterized in that, The multimodal semantic model construction module includes: The machine vision processing unit is used to extract features from the knowledge base graph and generate a structural topology graph. Triple extraction unit is used to construct an engineering object ontology knowledge graph based on dependency parsing; Incremental learning units are used to trigger online fine-tuning of the model when specifications change.

10. A method for analyzing multi-dimensional user needs during the engineering construction phase, characterized in that, Includes the following steps: A multimodal semantic model for the engineering construction phase is constructed. The multimodal semantic model integrates a multi-source knowledge base and extracts entity, attribute, and relation triplet description statements from the multi-source knowledge base text based on dependency parsing to construct an engineering object ontology knowledge graph. Through demand structuring guidance and intelligent parsing engine, the system dynamically identifies engineering stages and user roles, guides users to input textual requirements, and performs semantic completion and standardization processing on the user input based on the engineering object ontology knowledge graph. Based on a multi-dimensional rule base built from a multi-source knowledge base, conflicts in user needs are detected, and conflict impact analysis reports and coordination and handling solutions are generated. Based on the entity, attribute, and relationship triples obtained from the requirements analysis, the system automatically generates an ER diagram of the engineering object, an interactive prototype diagram, and a workflow diagram. Generate executable parameterized requirements documents and related diagrams to complete the closed-loop management of the project from requirements to design to verification.

Images