Action denormalization factor is different from voltage normalization factor

[max-schenke]

Apparently, the applied action is interpreted in terms of the supply voltage, but the state vector I get as output is interpreted in terms of the voltage limit that is specified for the motor.

E.g.: I apply a 4 quadrant converter at u_sup = 200 V.

action = 1 maps to 200 V applied voltage
action = -1 maps to -200 V applied voltage

I specify a motor with a voltage limit of 160 V

state output:
action = 1:
voltage = 200 V
normalized voltage (state output) = 200 V / 160 V = 1.25

or
action = -1:
voltage = 200 V
normalized voltage (state output) = - 200 V / 160 V = -1.25

Therefore, "normalized input" is not the same as "normalized state" because the first is referenced to the supply and the second is referenced to the motor.

If handled correctly, this is not necessarily wrong. But it does not simplify usage either.