| The automated infusion of therapeutic agents for control of cardiovascular variables has the potential to improve patient care, increase safety, and reduce healthcare costs. Grey Box Adaptive Control (GBAC) promises to address these issues by incorporating a priori known information and parameter constrained identification into the adaptive control strategy.; GBAC for automated drug delivery was tested on two single input single output systems (SISO): sodium nitroprusside (SNP) to lower and regulate mean arterial blood pressure (MAP) and dobutamine (DOB) to raise cardiac output (CO). The GBAC technology was advanced and tested in a multiple input multiple output system (MIMO) to simultaneously control MAP and CO using the concomitant infusion of SNP and DOB. To address issues relevant to tracking performance, stability, safety, and MIMO feasibility, GBAC was compared against standard black box adaptive control (BBAC) for both the SISO and MIMO cases using a modified benchmark cardiovascular model.; Tests for the MAP/SNP case compared the controllers for three types of simulated patients: healthy, hypertensive, and chronic heart failure. Tests for the CO/DOB case and for the combined SNP/DOB therapy concentrated on the chronic heart failure patient. Each case was further tested under a variety of disturbances including several noise and anesthetic levels, and during acute heart failure where appropriate. Statistical comparisons of several controller performance measures, concluded that GBAC, when compared to BBAC, enhanced steady state regulation and disturbance rejection, diminished variation of the controlled and manipulated variables, reduced the total amount of drug infused and increased reliability by maintaining the controlled variables within tolerances. GBAC also enhanced safety by reducing over and under shoots of the controlled variables, increasing safe regulation times and disturbance rejection, and decreasing controller bursting and on/off behavior. Evaluation of MIMO control indicated that GBAC was not only feasible, but also consistently stable and reliable.; In summary, this study has shown that GBAC shows great promise for the control of cardiovascular variables and other similar non-linear, time-varying systems. An automated infusor based on GBAC thus has the potential to positively impact the health care industry by increasing safety and reducing the cost of patient care. |
