An acausal modeling framework for automatically parallelized scientific machine learning (SciML) in Julia. A computer algebra system for integrated symbolics for physics-informed machine learning and automated transformations of differential equations
For differential equations there already are both methods for OD and DO, that deal with the fact that time is continuous.
But for infinite dimensional optimization problems the optimality criteria can also be derived before discretization.
This can lead to different optimal controls:
http://files.iam.metu.edu.tr/bulent_karasozen/abstracts/Hamdullah_sunum.pdf#page=10
For differential equations there already are both methods for OD and DO, that deal with the fact that time is continuous. But for infinite dimensional optimization problems the optimality criteria can also be derived before discretization. This can lead to different optimal controls: http://files.iam.metu.edu.tr/bulent_karasozen/abstracts/Hamdullah_sunum.pdf#page=10