The difference between KF (model fit) and KF (model mismatch) about kalmannet_tsp HOT 2 OPEN

Regarding the two questions I mentioned, perhaps we should use the one you mentioned in your paper: "In our numerical research, when introducing model uncertainty, we optimize the performance of the MB algorithm by carefully adjusting the covariance matrix, usually through grid search."? Where is this part reflected in the code? Or does this part of the work require me to manually adjust the code covariance matrix myself?

from kalmannet_tsp.

Or, how do I modify the parameters in the main code package and the architecture - # 1 code package when I need to draw images under certain information in my paper? I suspect it was my improper operation process. Finally, I also encountered a problem that I have been unable to solve:

from kalmannet_tsp.