issues
search
numagic
/
lumos
scalable accelerated optimal control
MIT License
16
stars
0
forks
source link
handle decisioni variable scaling manually
#66
Closed
yunlongxu-numagic
closed
2 years ago
yunlongxu-numagic
commented
2 years ago
previously the decision variable scaling is done by IPOPT, but now we take control of this ourselves to get better convergence and tolerance control.
this means the objective, constraints, their derivatives and the bounds on the decision variables all need to be scaled by ourselves.