Nonlinear fluid systems identified by random noise input.

The usual impedance measurement by Fourier analysis leads to undetectable errors when applied to nonlinear fluid systems where a nonlinear pressure drop p = K X V2 adds to the linear pressure-flow relationship, but processing series of impedance data with various input levels allows the measurement of such systems. This new method has been tested on simulations, nonlinear resistances, and intubation tubings. It gives accurate estimates of the nonlinear coefficient K together with the true linear impedance of the measured system. This way, the usual respiratory impedance measurement may be applied to intubated patients, as the respiratory impedance of the subject can be separated from the nonlinear behavior of the endotracheal tube or cannula.