基于规则的空管地空通话语音文本角色分类和文本分割算法

Rule-based Algorithm for Role Classification and Text Segmentation of Air Traffic Control Voice Text

摘要

针对空管语音识别应用中普遍存在的管制员与飞行员语音混合、角色识别和分割困难的问题，本文提出一种基于规则的地 空通话语音角色分类和通话文本分割算法，该算法设计了一套包含航空器呼号、数字组合等通用标签及逻辑符号的通话模 板体系，在模板侧采用抽象语法树对通话模板进行解析，在通话文本侧通过对识别文本先进行预处理，再通过顺序匹配算 法实现对混合语音文本的角色判定与自动分割。该方法有效地解决了管制员、飞行员语音的角色识别和混合对话分割问 题，为空管语音的语义理解、规范性分析及自动纠错等应用场景提供了一种技术途径。

Abstract

In response to the prevalent issues of mixed voice recognition, role identification, and segmentation difficulties between air traffic controllers and pilots in air traffic control (ATC) speech recognition applications, this paper proposes a rule-based algorithm for ground-air communication speech role classification and communication text segmentation. The algorithm designs a communication template system that includes universal labels such as aircraft call signs and numerical combinations, as well as logical symbols. On the template side, an abstract syntax tree is used to parse the communication templates. On the communication text side, the recognized text is preprocessed first, and then the role determination and automatic segmentation of mixed voice text are achieved through a sequential matching algorithm. This method effectively solves the problems of role identification and mixed dialogue segmentation between air traffic controllers and pilots, providing a technical approach for semantic understanding, normative analysis, and automatic error correction in ATC speech applications.

关键词

空管语音识别 / 角色分类 / 文本分割 / 规则 / 通话模板 / 匹配 / 抽象语法树

Key words

air traffic control speech recognition / role classification / text segmentation / rules / dialog template / matching / Abstract Syntax Tree

参考文献

[1] 郑诚, 唐思琪. 结合标签语义和对比学习的小样本文本分类[J/OL]. 小型微型计算机系统,1-14[2025-12-05]. https://doi.org/10.20009/j.cnki.21-1106/TP.2025-0319.

[2] 赵海燕, 曹杰, 陈庆奎, 等. 层次多标签文本分类方法[J]. 小型微型计算机系统, 2022, 43(04):673-683.DOI:10.20009/j.cnki.21-1106/TP.2021-0658.

[3] 崔海越, 李延玲. 基于BERT模型的多标签文本分类[J]. 高师理科学刊, 2025, 45(08):18-24.

[4] 孙晓瑜. 基于机器学习的文本分类与标签预测算法[J]. 吉林大学学报(信息科学版), 2025, 43(04):837-843.DOI:10.19292/j.cnki.jdxxp.2025.04.010.

[5] 贾桂敏, 申玉茜, 周熹龙. 基于层级规则树与特征融合的陆空通话规范性检验模型[J/OL]. 北京航空航天大学学报,1-15[2025-12-05]. https://doi.org/10.13700/j.bh.1001-5965.2025.0041.