Abstract: The power equipment standards aim to regulate design and manufacturing processes. However, the complexity and diverse sources of standard clauses make it challenging for frontline staff to select applicable terms. Traditional methods like stage feedback and expert reviews suffer from low efficiency and long cycles. Knowledge graphs, representing data relationships in a structured manner, facilitate the unified management of power equipment standard clauses. Uncovering implicit clause variance within existing standards is considered a knowledge graph completion problem. A method based on knowledge graph is proposed to identify clause variance, leveraging pre- trained language models to capture semantic associations between terms. The effectiveness of the proposed approach is validated through practical examples, providing robust support for the coordinated standardization of power equipment documents.