we only use the energies to get the temperatures, so why not do it directly?
leads to unnecessary computations in optimization model
But the most important reason is:
Their initialization! Within Dymola, the initialization for u is computed from the inital values for the temperatures! This is working in simulation with JModelica as well, but not with Optimization, which causes it to fail! And we do not want to compute the energies from our temperature measurements to be able to run the optimizations!
Some Reasons:
But the most important reason is: