An Adaptive Optimal Excitation Signal Control Method for Forced Oscillation Technique-Based Respiratory Impedance Measurement System

Nowadays the forced oscillation technique (FOT) for respiratory impedance measurement is increasingly used in clinical settings due to its simplicity, but current multifrequency measurement systems are unable to autonomously adjust the measurement frequency combinations based on clinical requirements, nor can they adaptively output the optimal excitation signals for different frequency combinations, limiting both flexibility and measurement quality. This study established a model for the FOT measurement system and analyzed the time-frequency domain response using Fourier analysis. Based on clinical studies of the FOT excitation signal, we first proposed a phase-genetic algorithm to search for the optimal phase combination that minimizes the peak-to-peak amplitude of pseudo-random signals, providing a phase reference for the control of the excitation signal. Then we proposed an optimal excitation signal control method based on the principle of linear superposition of signals and the concept of closed-loop control, which does not require specific model parameters of the system. It performs zero-phase initial superposition of signals with arbitrary frequency combinations to drive the loudspeaker and generate an output excitation signal. The excitation signal is sampled, decomposed, and estimated in the time-frequency domain to adjust the amplitude and phase of different frequency components of the drive signal, achieving optimal output for the measurement excitation signal. This method is highly adaptive to different input frequency combinations and system parameters. Experiments with various loudspeaker sizes and test frequencies verified the method’s feasibility and applicability, showing significant improvements in detection accuracy and repeatability.