Grey box adaptive control of cardiac output

Grey-box adaptive control (GBAC), i.e. the use of parameter-constrained identification with adaptive control, was compared against standard black-box adaptive control (BBAC) for the automated infusion of dobutamine (DOB) to raise and regulate cardiac output (CO). Using a benchmark cardiovascular model, tests compared the controllers for a chronic heart failure simulated patient under five different anaesthetics and five noise levels. Under quiescent conditions, when parameter constraints were not needed, there were no differences between the controllers. In the challenging cases, GBAC not only significantly (p<0.05) outperformed BBAC in almost all performance measures but demonstrated its robustness by successfully handling disturbances that repeatedly impaired BBAC. These results indicate that, when needed, GBAC rejected disturbances better, enhanced steady-state regulation and generated a smaller CO variation than BBAC