Open jkleiber opened 5 years ago
Current implementation is now to follow directives from the raspberry pi. The motion doesn't need to be super smart, so this has been moved to the pi. The Ackermann steering model is now implemented on the pi to find the desired steering angle. This needs to be tested
Description Outside of a special situation (Camera tracking the cone) the motion board needs to follow the navigational instructions of the main board
Action to take Implement some error-reducing system (like PID) that will keep the robot on the path, as desired by the main board. This may need to be more sophisticated than a straight PID to account for how the robot turns in a non-linear way. There are some motion equations that describe the robot's motion quite well - look up Ackermann steering for more information (these calculations are used in the kalman filter, and exist in that documentation as well)
Possible Solution Given the motion equations and the current navigational instructions, calculate what steering angle is needed to lock to the path. All of this code will exist on the motion board.