Overcoming the deadband in VESC motor control

[shulnak09]

Hi,

I am working on planning for one tenth scaled car that uses MPC to obtain velocity and steering angle as control input. The MPC works perfectly fine in simulation but when implementing on the F1-tenth scaled car, the wheel does not move especially at lower speeds say less than 0.5 m/s. So, even if my MPC outputs the control commands, the car is stalled at a place. I believe the the control command is unable to overcome the dead band torque in motors. Can anyone help me in this regard, if they have figured out how to accurately calibrate the vesc to overcome this static friction at lower control inputs?

Regards
Anshul