Abstract
In order to assess left ventricular support in heart failure patients, a lumped parameter model of a rotary blood pump and the cardiovascular system was used to investigate different control strategies. In this paper, a sensor-less control strategy for a left ventricular assist device (LVAD) based on pulsatile flow estimation has been presented, which referred to the crucial issue in the usefulness of LVADs is how to control the operation of the pump rotational speed to cater for cardiovascular system perturbations and changing metabolic demand for heart failure patients. A tracking control algorithm based on a robust sliding mode technique was developed to track the error difference between the reference pump flow and estimated mean pulsatile flow. A lumped parameter model of cardiovascular system in combination with the stable dynamical model of pulsatile flow estimation was used to evaluate the controller. The control algorithm was tested using constant and sinusoidal reference pump flow inputs under healthy and heart failure conditions. Tracking control was illustrated in the presence of modeling uncertainty and disturbance. Simulation results demonstrated that the control algorithm was able to track the reference input with minimal error in the presence of model uncertainty.