← 返回首頁Back to Home

外骨骼控制策略技術版圖Exoskeleton Control Strategy Landscape

X: 意圖感知前饋性  |  Y: 個體適應性X: Intent Prediction Lead Time  |  Y: Individual Adaptability

TIER 1 預定義驅動TIER 1 Pre-defined Drive
TIER 2 反應式感知TIER 2 Reactive Sensing
TIER 3 前饋預測TIER 3 Feedforward Prediction
意圖感知前饋性 →Intent Prediction Lead Time →
個體適應性 →Individual Adaptability →
事後補償Reactive
前饋預測Feedforward
固定族群Population-fixed
即時個人化Real-time Personalized
Gap
Tier 1 預定義驅動Tier 1 Pre-defined Drive
Tier 2 反應式感知Tier 2 Reactive Sensing
Tier 3 前饋預測Tier 3 Feedforward Prediction
ExoPulse (本研究)ExoPulse (This Study)
控制策略層級對照Control Strategy Tier Comparison
Tier 1
預定義驅動 ─ CPG、固定軌跡、能量最佳化。無使用者意圖感知。Predefined Drive ─ CPG, fixed trajectory, energy optimization. No user intent sensing.
Tier 2
反應式感知 ─ EMG 阻抗控制、sEMG 疲勞監測、pHRI 互動力分析。Reactive Sensing ─ EMG impedance control, sEMG fatigue monitoring, pHRI interaction force analysis.
Tier 3
前饋式意圖預測 ─ EMG-NMS 力矩、NMS+ML、EMG-DRL 共享控制、LSTM 軌跡預測。Feedforward Intent Prediction ─ EMG-NMS torque, NMS+ML, EMG-DRL shared control, LSTM trajectory prediction.
ExoPulse
意圖預測 × 個體適應性最高 ─ 數位孿生 + DRL 雙代理 + Wrapper 臨床介面 + 腦機介面與數位永生。Intent Prediction × Highest Individual Adaptability ─ Digital Twin + DRL dual-agent + Wrapper clinical interface + BCI & Digital Immortality.
ExoPulse 閉迴路生態ExoPulse Closed-Loop Lifecycle
步態感測Sensing → 指標分析Analysis → 策略建議Strategy → 執行訓練Training → 再感測Re-Sense ⟳
ExoPulse Patient Cloud × ExoPulse Studio