Acta Technica Jaurinensis - Vol. 8. No. 2. (2015.)

This paper discusses a state space estimation for an electromechanical actuator valve using extended Kalman filter (EKF). A proposed actuator model includes a Tustin’s friction model with strong nonlinearities, hence it represents an accurate model for describing the friction phenomenon in an electromechanical actuator valve. At first the state equations of the model were converted into a time-discrete. Then the EKF is used for estimation of three states, namely the motor current, valve angular velocity and position. Both the actuator model and the EKF algorithm were implemented in MATLAB®. The measurements for input data in the EKF are performed in National Instruments CompactRIO system. Outline of our examination should be to take a conclusion, how much EKF is capable of accurate estimation of the desired states. Furthermore we will use this method for further developing the fault detection in the electromechanical actuator valve.