Hello,
I have a model comprising of several non linear equations. The aim is to instantiate these models several times to form a system.  The high non linearity of the equations in the model, allows it to often cause convergence issues to some mathematical violations/complexities. Is there any way such that start values for target variables can be provided at every instantiation rather than the variables taking the same start values for every instantiation.
Somewhat like linear interpolation but from outside the model?

https://github.com/FOSSEE/OMChemSim

A glimpse of the model can be viewed here. Its a distillation column comprising of trays, reboilers and condensers.