There should be an easy way to initialize the induction machine at its nominal point.

[ceraolo]

The only induction machine example proposed in MSL initialises it at no load.

I tried to create a model in which it intialises at its nomial point. Probably such an example could be useful as a starting point for initialisation different from zero in other use cases.

As a first attempt to initialise the machine at its nominal point I created a model having as diagram the following:

The applied voltage is the nominal value for the machine:

Parameters inside the machine are kept unchanged (so they get the hard-coded internal values) and the torque has as parameters the following ones:

The only initial condition I set is the inertia initial speed, that I set equal to the nominal value, i.e. 1440.45 rpm.
The result in terms of torques and rotational speed are as follows:

We see here a significant transient, which is obviously due to the state variables different from speeds are not set to their steady-state values.

But... What are these other state variables?
Dymola says it misses initial values for the following four continuous states:
idq_rr[1..2]
idq_sr[1..2]

that is a strong indications that these are taken as states, which is very reasonable.

So, the idea, before thinking to how to implement automatic initialisation for the machine is to first manually set the values of these values to their steady state values.
The trends of these currents are as follows:

them, I read their final values, i,e,:
idq_rr={77.1 -104.1}
idq_sr={-43.6 136.0}

set them as initial values, and re-run, hoping to find the system staying in steady-state.
To my surprise, if I set these values as initial values for the corresponding quantities, I still get a large transient:

This indicates that my approach in trying to have the machine in steady.state is faulty.

So my questions are:

1. What is the reason for which setting all the significant state variables (w, idq_rr, idq_sr) to their final values (those at t=1s) did not cause the simulation to start in steady-state?
2. Maybe an example can be added to MSL in which the machine is initialised to its nominal point? This can be simply a modification of the existing IMC_Initialize model. This example could help users understand how to manage such situations and similar ones.
3. Maybe is a good idea to implement code for steady-state initialisation in such a way that, if the user checks a checkBox saying "initialise in steady-state" all the derivatived of state variables are set to zero? Several models exist in the MSL proposing this option, for instance the PID controller, which results very useful in several cases.