estimateSegmentTimes about mav_trajectory_generation HOT 4 CLOSED

If I remember correctly the problem with t=0.5*v_max is that it can lead to high accelerations in case of short segment times. The new heuristic was developed by my student nfabian, that's why the constant is named after him.

from mav_trajectory_generation.

Hi Marcelino,

this function is only a heuristic that makes an overly cautions guess on the trajectory times. Usually, this results in quite slow but feasible trajectories. Do you know what the assumption for this guess is @helenol ?
t = 0.5*v_max is also a valid heuristic which is usually feasible.

Note that for rest-to-rest trajectories you can even compute a minimum time for a feasible trajectory [Müller et al.].

from mav_trajectory_generation.

@burrimi, would fabian have a documentation on his heuristics?

Thanks!

from mav_trajectory_generation.

Now replaced (optionally) with trapezoidal time allocation, which is quite a bit more sensical. And no, no documentation exists... ;)

from mav_trajectory_generation.