Incorrect result of a thyristor bridge with a star/star transformer using the multi-time-step solver

aesim-tech / simba-project

Roadmap & issue tracking

7 stars 0 forks source link

Incorrect result of a thyristor bridge with a star/star transformer using the multi-time-step solver #324

Closed jcornau closed 1 year ago

jcornau commented 1 year ago

Describe the bug A simulation with a thyristor bridge and a star/star transformer is giving wrong results at specific time points when using the multi-time-step solver.

It seams some commutations of the diods are not expected. The differences can also vary when changing the circuit (using a different grid frequency) or the simulation settings (different minimal time step).

Screenshots

Desktop

OS: Windows 11
Language: Any

SIMBA Version

Version 2023.08 & 2023.05

manu-aesim commented 1 year ago

@jcornau , I think it is now fixed and the test is added to the CI pipeline. Can you please double check the original file causing the issue with the latest beta release ?

jcornau commented 1 year ago

@manu-aesim I checked with the original model and the problem is fixed in both cases:

predictive time step and Sampling time activated,
fixed time step solver.