Prediction the Position of Artificial Heart Motor Actuator Using Neural Network Time Series

Cardiovascular diseases (CVDs) are of the most prevalent diseases around the world and the total artificial heart (TAH) is taken as a temporary solution to survive the patient. In the present study, a special model of a TAH that fabricated at Aachen University,Germany, has been investigated and the position of the actuator of the pump unit of the TAH has been predicted with the help of the neural network time series. Identification of the actuator’s position is important, because the electric current entering the TAH pump willvary depending on the position of the actuator. For this purpose, first, the dynamic of system and state space equations were derived. Hence, an appropriate controller was designed for the system, using the sliding mode method. In the next step, the data of the first fiveperiods of the actuator reciprocating motion with 80 data in each period, were collected. Afterwards, the actuator’s path was estimated using MATLAB software and neural network time series. The results indicated that the reliability of the constructed models increasedthe use of artificial neural network (ANN) in this field.

Keywords:

Artificial Heart , Sliding Mode , Forecasting , Time Series