Datasheet
The interval (t
2
-t
1
) is the start-up time, and it is supposed to be large enough to allow the estimation algorithm to reach
the complete alignment with the real phase of the system.
At t = t
2
, the phase estimation process is supposed to be aligned. At this point a reference system change is performed:
from the arbitrary (d
a
, q
a
) reference to the (d, q) reference based on the estimated phase
est
.
The system reference change is performed as follows:
The current references in the (d
a
, q
a
) system, id
ref
=i
start
and iq
ref
=0, are projected in the fixed system (, ), to
compute the instantaneous current components; the same is done with the integral parts of the current PI
controllers, which are the mean voltages required to obtain those currents; in this way we obtain i
, i
, v
, v
.
The phase is updated to
est
.
The and components obtained before are projected into the new reference system (d, q), giving the new
current reference values i
d0
, i
q0
and the new PI integral memories v
d0
, v
q0
; the speed PI integral memory is
loaded with the q current reference.
After t > t
2
, the normal control is performed, based on the estimated phase
est
; the speed reference is increased with the
classical ramp; the i
d
current reference is decreased with a ramp, till it reaches the value zero at the moment t
3
; then it is
maintained to zero; the i
q
current reference is obtained as output of the speed PI controller.