An aviation speech recognition method based on position information assistance, medium and device

By combining real-time aircraft location information and personalized acoustic feature vectors, dynamically associating with a preset location lexicon and performing constrained secondary recognition, the problem of misidentification of technical terms in aviation communications is solved, achieving highly accurate aviation speech recognition and ensuring flight safety.

CN122417031APending Publication Date: 2026-07-17HANGKE TECH DEV +1
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202610859620.3
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-06-15
Publication Date
2026-07-17

AI Technical Summary

Technical Problem

Existing aviation communication voice recognition systems have problems recognizing aviation terminology, which can cause pilots to receive incorrect position instructions and affect flight safety.

Method used

By dynamically associating pre-defined location terms with real-time aircraft location information, and combining the recognition of pre-defined location guide words and candidate location words, the accuracy of professional terminology recognition is improved by employing constrained secondary recognition and adaptive adjustment of personalized acoustic feature vectors.

Benefits of technology

It significantly improves the accuracy of identifying technical terms in aviation communications, reduces the misidentification rate caused by interference factors such as accents and speech speed, ensures that pilots receive accurate position instructions, and provides efficient and reliable aviation communication assistance.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122417031A_ABST
    Figure CN122417031A_ABST
Patent Text Reader

Abstract

本发明涉及航空通讯辅助技术领域,特别是涉及一种基于位置信息辅助的航空语音识别方法、介质及设备。该方法通过飞机实时位置确定当前地理区域并关联预设位置词库,对无线电语音进行初步识别后,通过预设位置引导词定位候选位置词,判断其是否存在于词库中;若不存在则进行约束性二次识别,将候选词集限定为词库内词汇以修正错误。本发明结合位置上下文动态缩小识别范围,通过二次识别机制提升专业术语识别准确性和鲁棒性,有效降低口音、语速等干扰导致的误识别率,为飞行员提供可靠指令文本辅助。
Need to check novelty before this filing date? Find Prior Art