search

helgeanl / GP-MPC

MPC with Gaussian Process
MIT License
193 stars 60 forks source link

problem with predict_compare function #5

Open Fangwq opened 4 years ago

Fangwq commented 4 years ago

Hi, thanks for sharing the code. I have a question for the function predict_compare in file gp_class.py. In the following figure, when I use if feedback:, the code goes wrong(At t = 0.00759683, mxstep steps taken before reaching tout.). How to solve this problem? Thank you very much! image

helgeanl commented 4 years ago

It has been a while since I looked at this code, so can't say for sure, but the problem is probably that the integrator is diverging and unable to find a solution. I used the solver from Sundails, with CVODES for ODEs and IDEAS for DAEs, this is automatically set depending on what model type you specify. You can see a more in depth discussion about this in this thread. In general you need to provide suitable control matrices if you are using the LQR feedback, or find an input time series u that don't cause your system to diverge in the open loop simulation. I can't say much more without more context.