Design of a Wearable Physiological Signal Feedback Control System for Personalized Health Regulation

Abstract

Wearable closed-loop systems for personalized health regulation face significant challenges in dynamically adapting control parameters to account for individual physiological variability. To address these limitations, this study designs, implements, and rigorously evaluates an advanced feedback control system based on the continuous monitoring of photoplethysmography (PPG) and electrodermal activity (EDA) signals. The proposed system seamlessly integrates real-time physiological signal acquisition, robust adaptive filtering techniques, and a sophisticated proportional-integral-derivative (PID) controller equipped with dynamic gain scheduling for highly personalized health regulation. A comprehensive mixed-methods approach, combining quantitative signal processing techniques and comparative performance analysis, was systematically employed. The overall system architecture was extensively evaluated using three prominent, publicly available physiological databases: PhysioNet's MIMIC II (n=25 subjects), WESAD (n=15 subjects), and the UCI TEPHRA dataset (n=20 subjects), collectively comprising over 200 hours of diverse multimodal data. The experimental results conclusively demonstrate that the proposed adaptive control algorithm achieves an impressive signal-to-noise ratio of 28.4 dB, representing a substantial improvement compared to the 21.2 dB observed in conventional fixed-gain methods. Furthermore, the system successfully tracked dynamic physiological setpoints with a minimal steady-state error of 2.3% and consistently maintained robust control stability across various simulated physiological states. The average response time for effective feedback regulation was recorded at a highly efficient 1.8 seconds. Ultimately, this study contributes a thoroughly validated system design framework and establishes an open algorithmic baseline for next-generation personalized biofeedback applications, offering profound insights for the future integration of wearable technologies in proactive and preventive health management.