Computer simulation of sensorless fuzzy control of a rotary blood pump to assure normal physiology.

Abstract

Rotary blood pumps have been considered effective permanently implantable devices. However, control of such pumps is quite complicated. Sensorless control of pump flow is required because no invasive flow or pressure sensors are wanted. Whereas insufficient pump output can cause underperfusion and should be avoided, overpumping may cause ventricular collapse and must be prevented. An intelligent physiologic control algorithm is highly desirable to reach optimal pump output based on physiologic requirements. We present an intelligent physiologic control mechanism for the blood pump allowing it to achieve normal physiology. Sensorless control of pump flow is gained by analysis of the electric motor current and speed. The required pump output flow is chosen based on heart rate, and an intelligent fuzzy logic based control mechanism is developed to adjust the motor input so that the pump output can reach required flow while also preventing the occurrence of ventricular suction or cannular collapse. Computer simulation was carried out, and the results indicate that the proposed algorithms can achieve required pump flow to obtain normal physiology, whereas overpumping can be prevented to provide safe operation.