02-05-2012, 11:13 AM
CONTROL OF AN ELECTROMECHANICAL CAMLESS VALVE ACTUATOR
20113301295224.pdf (Size: 865.85 KB / Downloads: 102)
INTRODUCTION
In recent years camless engine has caught much attention in
the automotive industry. Camless valvetrain offers programmable
valve motion control capability. However, it also introduces
valvetrain control issues. There are mainly two types of camless
actuators, electrohydraulic valve (EHV) [4][7] and
electromechanical valve (EMV) [1][2][3][6][8][9][12] actuators.
This paper deals with the EMV type of actuator.
MODELING ANALYSIS
An EMV system consists of two opposing electromagnets, an
armature, two springs and an engine valve. The armature moves
between the two magnets. When neither magnet is energized, the
armature is held at the mid-point of the two magnets by the two
springs located on either side of the armature. This system is used
to control the motion of the engine valve. The engine valve is
then in turn used to control the flow of air into and out of a
combustion engine cylinder.
CONTROL DESIGN
When the valve-closing event starts, the lower solenoid coil
in Figure 3 is deactivated, and the valve moves up towards its
seating position by the mechanical spring force. An EMV
actuator works according to the spring-mass pendulum principle,
which means that the system follows its own natural oscillation
frequency, and external electromagnetic force is only needed for
overcoming the friction loss. The electromagnetic actuator is only
effective in a relatively short range closing to the seating position,
and so it is not efficient in the sense of energy consumption to
apply closed-loop control when the valve is still far away from the
seating position. However, previous analysis in section 2.4 shows
that the system goes unstable as the engine valve moves to the
region within one-third of the total lift.
CONCLUSIONS
Our mathematical modeling analysis reveals that a camless
electromechanical valve actuator becomes unstable, as the
distance of the engine valve to its seating position is less than
roughly one third of the total lift, regardless of spring rate or
electromagnetic coil turns. A linear model was obtained based on
the derived model structure and system identification test result
for control system design. An LQ optimal control is designed and
implemented on the hardware system. The experimental results
have shown consistent closed-loop system response.