A Data-Driven Gain Tuning Method for Automatic Hovering Control of Multicopters Via Just-In-Time Modeling

摘要

This study develops a new automatic hovering controlmethod based on just-in-time modeling for a multicopter. Especially, themain aim is to compute gains of a feedback control law such that themulticopter hovers at a desired height and at a desired time without overshoot/undershoot. First, a database that contains various hovering data isconstructed, and then the proposed method computes gains for a query inputfrom the database. From simulation results, it turns out that the multicopterachieves control purposes, and hence the new method is effective.