基于语音识别灯光智能控制系统研究  

Research on Intelligent Light Control System Based on Speech Recognition

作　　者：牛瑞坤 李冬辉[1] 

机构地区：[1]天津大学电气与自动化工程学院,天津 [2]金陵科技学院机电工程学院,江苏 南京

出　　处：《动力系统与控制》2024年第4期153-159,共7页Dynamical Systems and Control

摘　　要：近年来,虽然汽车技术在不断发展,但控制类汽车电子技术的发展却进步较为缓慢,从而导致国内汽车产业在电子技术方面存在一些短板。本研究旨在设计并实现一套基于语音识别的汽车灯光控制系统,以提高驾驶的安全性和操作的便捷性。同时,现有的汽车电子控制系统多依赖于国外的技术,这不仅限制了国内技术的发展,也影响了汽车产业的自主创新能力。本文对语音识别系统进行设计研究,并对比了普通话和方言在语音识别中的准确率。实验结果显示,普通话识别模型准确率高,四川话、广东话和河南话的准确率分别为73.8%、76.2%和59.5%。In recent years, although automotive technology has been continuously developing, the development of control automotive electronic technology has not progressed relatively slowly, resulting in some shortcomings in the domestic automotive industry in terms of electronic technology. This study aims to design and implement a car lighting control system based on speech recognition to improve driving safety and operational convenience. Meanwhile, existing automotive electronic control systems rely heavily on foreign technology, which not only limits the development of domestic technology but also affects the independent innovation capability of the automotive industry. This article conducts design research on speech recognition systems and compares the accuracy of Mandarin and dialects in speech recognition. The experimental results show that the Mandarin recognition model has high accuracy, with accuracy rates of 73.8%, 76.2%, and 59.5% for Sichuan dialect, Cantonese, and Henan dialect, respectively.

关 键 词：语音识别 灯光控制 方言识别 

分 类 号：F42[经济管理—产业经济]