Data-driven Model Predictive Controller Design for Left Ventricular Assist Devices

Patients with chronic heart failure usually implant an assist device to help the ventricle in the process of blood pumping throughout the blood circulation system. In this paper, the cardiovascular system's state-space representation with the left ventricular assist device (LVAD) is exploited. After that, a data-driven model predictive approach for left ventricular assist device (LVAD) is presented. The goal is to be able to control the LVAD flow in a patient who uses this device to experience normal life and improve the life quality. Control of these devices is such a challenging phenomenon because if the control system fails, ventricular suction could happen. The proposed approach addresses the challenges effectively and controls the system's variables to provide the cardiac output in an acceptable range. Simulation results showed that the controller could track the flow of a healthy heart and provide our aims.