The use of an electromechanical valve actuator (EMVA) formed by two magnets and two balanced springs is a promising tool to implement innovative engine management strategies. This actuator needs to be properly controlled to reduce impact velocities during engine valve operations, but the use of a position sensor for each valve is not possible for cost reasons. It is therefore essential to find sensorless solutions based on increasingly predictive models of such a mechatronic actuator. To address this task, in this paper, we present an in-depth lumped parameter model of an EMVA based on a hybrid analytical-finite-element method (FEM) approach. The idea is to develop a model of EMVA embedding the well-known predictive behavior of FEM models. All FEM data are then fitted to a smooth curve that renders unknown magnetic quantities in analytical form. In this regard, we select a single-wise function that is able to describe global magnetic quantities as the flux linkage and force both for linear and saturation working regions of the materials. The model intrinsically describes all mutual effects between two magnets. The goodness of the dynamic behavior of the model is finally tested on a series of transient FEM simulations of the actuator in different working conditions.
Published in:
Mechatronics, IEEE/ASME Transactions on
(Volume:13
,
Issue:
6
)
Date of Publication: Dec. 2008