Open sshane opened 2 years ago
Have a net predict an accurate accel request to follow the desired plan, specific to how Toyota responds to acceleration commands.
Input (training) -> [current aEgo, future aEgo] Input (inference) -> [current aDesired, lag compensated aDesired] Output -> [current accel cmd]
Have a net predict an accurate accel request to follow the desired plan, specific to how Toyota responds to acceleration commands.
Input (training) -> [current aEgo, future aEgo] Input (inference) -> [current aDesired, lag compensated aDesired] Output -> [current accel cmd]