US 11,987,132 B2
Linear time varying model predictive torque control
Yue-Yun Wang, Troy, MI (US); Lei Hao, Troy, MI (US); and Suresh Gopalakrishnan, Troy, MI (US)
Assigned to GM GLOBAL TECHNOLOGY OPERATIONS LLC, Detroit, MI (US)
Filed by GM Global Technology Operations LLC, Detroit, MI (US)
Filed on Jan. 11, 2021, as Appl. No. 17/145,598.
Prior Publication US 2022/0219548 A1, Jul. 14, 2022
Int. Cl. B60L 15/20 (2006.01); B60L 50/51 (2019.01)
CPC B60L 15/20 (2013.01) [B60L 50/51 (2019.02); B60L 2210/42 (2013.01); B60L 2240/423 (2013.01); B60L 2240/427 (2013.01); B60L 2240/429 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A system for torque control of an electric motor in a motor vehicle, the system comprising:
a power inverter that delivers a current to the electric motor to regulate the torque of the electric motor; and
a model predictive control (MPC) module that sends a three-phase voltage to the power inverter to control the operation of the power inverter, wherein the MPC module minimizes a cost function to find optimal control motor voltages, and wherein the MPC module includes:
a deep learning neural network that generates desired current commands based on a desired motor torque, motor speed, battery voltage, and motor temperature;
an inner loop that controls the current to the electric motor based on the desired current commands generated by the deep learning neural network; and
an outer loop that provides torque feedback control to the electric motor.