US 11,833,907 B2
Vehicle powertrain system with machine learning controller
Mohak Prafulkumar Bhimani, Sunnyvale, CA (US); Elnaz Vahedforough, San Jose, CA (US); and Mikhail Tetelman, Los Gatos, CA (US)
Assigned to VOLKSWAGEN AKTIENGESELLSCHAFT, Wolfsburg (DE)
Filed by VOLKSWAGEN AKTIENGESELLSCHAFT, Wolfsburg (DE)
Filed on Dec. 29, 2020, as Appl. No. 17/137,106.
Prior Publication US 2022/0203842 A1, Jun. 30, 2022
Int. Cl. B60L 15/00 (2006.01); B60L 15/20 (2006.01); G05B 13/02 (2006.01); B60K 6/28 (2007.10)
CPC B60L 15/20 (2013.01) [G05B 13/0265 (2013.01); B60K 6/28 (2013.01); B60Y 2200/92 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A method, comprising:
receiving a power request for a vehicle, wherein:
the vehicle comprises one or more power sources; and
each power source of the one or more power sources is configured to provide power for moving the vehicle;
obtaining perception data generated by a perception system of the vehicle;
determining, by a powertrain supervisor, a power distribution based on the power request, a set of perception data, a vehicle state and a machine learning model, wherein:
a first portion of the set of perception data is associated with a first distance interval from the vehicle that is based on a set of ranges of a set of sensor devices;
the power request and the set of perception data are provided as an input to the machine learning model; and
the power distribution is generated by the machine learning model; and
providing the power distribution to a powertrain controller, wherein:
the powertrain controller is configured to control the one or more power sources for the vehicle based on the power distribution; and
the powertrain supervisor is communicatively coupled to the perception system.