本体驱动非合规施工方案建模与隐患推理预警

Ontology-driven modeling and hazard early warning for non-compliant construction schemes

摘要

施工方案的编制质量直接影响施工安全、质量与效率。传统的施工方案编制和审查主要依赖专家，缺乏统一的模板，且流程费时费力、效率低下。为解决这一问题，基于本体理论，利用Protégé本体构建工具和“七步法”构建了施工方案隐患本体知识库，明确定义了类及其层次结构、对象属性、数据属性和约束条件。通过内置的HermiT推理机进行一致性检查，确保本体知识库的正确性。同时基于SWRL规则语言构建了隐患规则库，并通过Drools推理引擎对从施工方案文本中提取的关键信息进行推理，验证了规则库的可靠性与可行性。最后，通过案例验证推理功能，实现了施工方案隐患的自动预警。该方法有效提高了施工方案审查的效率与准确性，有助于发现施工方案潜在的安全与质量风险，减少事故隐患的发生。

Abstract

The quality of construction scheme preparation directly impacts construction safety, quality, and efficiency. Traditional construction scheme compilation and review heavily rely on experts, lack standardized templates, and often involve time-consuming and inefficient processes. To address these issues, an ontology-based knowledge base was developed for construction scheme hazards using the Protégé ontology construction tool and the “Seven-Step Method.” We clearly define classes and their hierarchical structures, object properties, data properties, and constraints, and create instances while ensuring consistency with the built-in HermiT reasoner. A hazard rule base is then built using Semantic Web Rule Language(SWRL), with key information extracted from construction scheme documents being reasoned through the Drools inference engine, verifying the rule base's reliability and feasibility. Finally, case studies validate the reasoning function and achieve automatic hazard early warning for construction schemes. This approach significantly improves the efficiency and accuracy of construction scheme reviews, and identifies potential safety and quality risks, reducing the occurrence of accidents.

关键词

本体 / 施工方案 / 隐患预警 / SWRL / 推理

Key words

ontology / construction scheme / hazard early warning / SWRL / reasoning

Author summay

谭柯鑫(2000—), 女, 硕士研究生, 主要从事基于深度学习的施工组织文档处理研究。E-mail: [e-mail]

基金资助

中国长江三峡集团有限公司科研项目(202103551)