Ontology Construction and Logical Consistency Reasoning for UHV Grid Operation Texts

Abstract

The operation of Ultra-High Voltage (UHV) power grids generates 
massive quantities of unstructured textual data, primarily in the form of dispatch logs, maintenance manuals, and accident reports. While these documents contain critical information regarding system states and operational procedures, their unstructured nature hinders automated processing and intelligent decision support. This paper proposes a comprehensive framework for ontology construction and logical consistency reasoning specifically tailored for UHV grid operation texts. By transforming unstructured textual data into a structured knowledge base, we aim to enhance the semantic interoperability of power system data. The methodology involves a domain-specific ontology construction phase that integrates the Common Information Model with linguistic characteristics of operation logs, followed by a semantic reasoning mechanism designed to detect logical inconsistencies in operational commands. We employ deep learning techniques for entity and relation extraction to populate the ontology, subsequently applying description logic rules to validate operational sequences against safety regulations. The experimental results demonstrate that the proposed framework significantly improves the accuracy of identifying unsafe operational directives and conflicting equipment states compared to traditional keyword-based approaches. This research provides a theoretical foundation and practical tools for realizing cognitive intelligence in modern smart grids.