Scholarlink

<p>Robot-assisted vocal training integrates technology with therapeutic techniques to enhance vocal performance and rehabilitation. Dynamic feedback control strategies are vital in such systems to adaptively support users in real-time based on their vocal and physical responses. However, existing methods often suffer from limitations in adaptability and fail to respond effectively to individual variations in muscle tension, pitch modulation, and posture during vocalization. To address these challenges, this study proposes an Adaptive Impedance Control (AIC) framework that continuously adjusts robotic resistance and assistance in response to real-time feedback. AIC enables the robot to provide variable support levels by sensing the user’s biomechanical and acoustic parameters, ensuring personalized and responsive interaction. Key metrics recorded include muscle activation level (reduced EMG RMS from 1.75 to 1.35 mV), pitch accuracy (error reduced from 15.2 to 4.2&#xa0;Hz), fatigue index (dropped from 0.71 to 0.4), and improved vocal loudness (from 68.2 to 74.5 dB). Additionally, impedance gain adjustment (ΔK) was dynamically modulated (42&#xa0;N/m to 18&#xa0;N/m), robotic assistance force decreased (from 5.8 to 3.1&#xa0;N), and postural deviation was reduced significantly (14.1° to 4.6°). User satisfaction also improved with feedback scores rising from 6.5 to 9.3. This confirms that AIC-based dynamic feedback control significantly contributes to safer, more effective, and individualized vocal training protocols.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Dynamic feedback control strategies in robot-assisted vocal training

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年6月23日

查看全文
  • Bei Xu

摘要

Robot-assisted vocal training integrates technology with therapeutic techniques to enhance vocal performance and rehabilitation. Dynamic feedback control strategies are vital in such systems to adaptively support users in real-time based on their vocal and physical responses. However, existing methods often suffer from limitations in adaptability and fail to respond effectively to individual variations in muscle tension, pitch modulation, and posture during vocalization. To address these challenges, this study proposes an Adaptive Impedance Control (AIC) framework that continuously adjusts robotic resistance and assistance in response to real-time feedback. AIC enables the robot to provide variable support levels by sensing the user’s biomechanical and acoustic parameters, ensuring personalized and responsive interaction. Key metrics recorded include muscle activation level (reduced EMG RMS from 1.75 to 1.35 mV), pitch accuracy (error reduced from 15.2 to 4.2 Hz), fatigue index (dropped from 0.71 to 0.4), and improved vocal loudness (from 68.2 to 74.5 dB). Additionally, impedance gain adjustment (ΔK) was dynamically modulated (42 N/m to 18 N/m), robotic assistance force decreased (from 5.8 to 3.1 N), and postural deviation was reduced significantly (14.1° to 4.6°). User satisfaction also improved with feedback scores rising from 6.5 to 9.3. This confirms that AIC-based dynamic feedback control significantly contributes to safer, more effective, and individualized vocal training protocols.