US 11,693,373 B2
Systems and methods for robust learning-based control during forward and landing flight under uncertain conditions
Guanya Shi, Pasadena, CA (US); Xichen Shi, Sunnyvale, CA (US); Michael O'Connell, Pasadena, CA (US); Animashree Anandkumar, Pasadena, CA (US); Yisong Yue, Glendale, CA (US); and Soon-Jo Chung, La Cañada, CA (US)
Assigned to California Institute of Technology, Pasadena, CA (US)
Filed by California Institute of Technology, Pasadena, CA (US)
Filed on Dec. 10, 2019, as Appl. No. 16/709,775.
Claims priority of provisional application 62/777,646, filed on Dec. 10, 2018.
Prior Publication US 2020/0183339 A1, Jun. 11, 2020
Int. Cl. G05B 13/02 (2006.01); G06N 3/04 (2023.01); G06N 3/084 (2023.01)
CPC G05B 13/027 (2013.01) [G06N 3/04 (2013.01); G06N 3/084 (2013.01)] 18 Claims
OG exemplary drawing
 
1. A method for training an adaptive controller, the method comprising:
receiving a set of training data comprising a plurality of training samples, each training sample comprising a state and a true uncertain effect value, wherein the state comprises vehicle dynamics data;
computing, for each training sample, an uncertain effect value based on the state;
computing, for each training sample, a set of one or more losses based on the true uncertain effect value and the computed uncertain effect value, wherein the set of one or more losses comprises a position tracking error; and
updating the adaptive controller based on the computed set of losses,
wherein the true uncertain effect value is a disturbance force caused by at least one of the group consisting of ground effects and wind conditions.