Adaptive Control for a Two-Compartment Respiratory System

Respiratory failure is a typical clinical issue which needs immediate help with mechanical ventilation while the hidden reason is recognized and treated. Mechanical ventilation can damage the lungs if the applied pressure is too high. It is desirable to provide the desired blood levels of CO2 and oxygen with limited pressure to avoid causing the lung injury. This paper uses the adaptive inverse dynamics control technique to control a two-compartment modelled respiratory system. Based on the nonlinear respiratory model and desired respiratory pressures, the adaptive inverse dynamics control scheme consisting of a control law and an adaptation law is then applied. The control law has the structure of the two-compartment inverse dynamical model but uses estimates of the dynamics parameters in the computation of pressure applied to the lungs. The adaptation law uses the tracking error to compute the parameter estimates for the control law, stops updating a given parameter when it reaches its known bounds, and resumes updating as soon as the corresponding derivative changes sign. Computer simulations to evaluate the control technique were conducted. Our results indicate that the tracking errors can be improved if the parameter values associated with the adaptation law are properly chosen, and the performance is also robust despite relatively large deviations in the initial estimates of the system parameters.