engleski

State Estimation of an Electronic Throttle Body

Electronic throttle body is a device used in cars to regulate air inflow into the motor engine. Its good behavior is crucial for the superimposed engine speed control system. However, electronic throttle body is a highly nonlinear system. Moreover, the only measured variable is the throttle valve position, and it is measured by a cheap potentiometer of low resolution. In this paper we apply Extended and Unscented Kalman Filters (EKF and UKF) for the throttle state estimations with two goals in mind. Firstly, availability of the internal states enables application of an advanced control strategy, e.g. Model Predictive Control. Secondly, state estimator also filters the valve position measurement signal, which is originally very noised and not suitable for control signal calculation. Both, EKF and UKF were implemented and tested by simulations and experiments on an electronic throttle body. While simulation does not show any advantage of the UKF, the experimental results show that UKF exhibits better performances.

Extended and Unscented Kalman Filter (EKF and UKF); electronic throttle; friction and limp-home nonlinearity

nije evidentirano