Currently the estimated position sensor in the mobilator manager appears to estimate the mobilator position by simply integrating the command velocity.
This tends to produce very inaccurate results on wheel-driven mobilators, since it does not model any effects from the starting wheel direction or any less-than-perfect performance of the PID loops.
It would be better if the estimated position model relied on encoder feedback from the wagon units rather than extrapolating from the command velocity. (When using wagon unit types that have encoder feedback, obviously this is not possible when using e.g. boat thrusters.)
Please sign in to leave a comment.