Abstract
High-tech mechanical ventilators are engineered to deliver precise and consistent airflow, which is critical for effective respiratory therapy. This study evaluates flow control performance in a custom-built electro-pneumatic ventilator prototype, comparing Proportional-Integral-Derivative (PID) control with Fuzzy Logic Control (FLC) through real-time experiments on a test-lung platform to assess accuracy and adaptability under dynamic conditions. A laboratory based experimental study was conducted under laboratory conditions, using a test lung simulator and real-time flow data acquisition. The analysis included time-domain performance metrics and statistical validation through Bland-Altman analysis. Results indicate that both controllers meet the accuracy thresholds expected in commercial systems. However, the fuzzy logic controller exhibited narrower limits of agreement and lower standard deviation, indicating greater consistency. While PID control responded faster, with a settling time between 0.32 and 0.43 seconds, FLC achieved superior performance in high-demand scenarios, delivering an entire volume of 900 mL. Stability analysis using the Jury Test and Nyquist criteria confirmed that both systems are dynamically stable. Notably, the FLC curve in the Nyquist plot remained farther from the critical point (-1, 0j), indicating enhanced robustness against disturbances. These findings suggest that FLC may offer a reliable alternative to PID in nonlinear ventilation scenarios, particularly in resource-constrained environments seeking technological autonomy.