# Follower PID Tuning

You are on the follower pid tuning step
You are here

This is the last tuning step. Here, you will tune the closed loop feedback controller. This should improve your path following dramatically.

There are two alternative opmodes for this process, BackAndForth or FollowerPIDTuner.

It is recommended to run the BackAndForth opmode to get a set of coarse PID gains and then run the same process using FollowerPIDTuner for further fine tuning. The BackAndForth opmode simply travels in a straight line back and forth. It is easy to see major errors in translation and heading PID accumulate here.

The FollowerPIDTuner will follow a large square. The bot will turn counterclockwise on each corner. If your heading is off, your entire square is destroyed and you have to keep resetting the bot. It's quite arduous. Use the BackAndForth opmode for initial tunings and then continue to further fine tune with FollowerPIDTuner.

  1. Run the BackAndForth opdmode via the RC.

  2. Then, connect to the RC phone's wifi network. The password to the network is located in the Program and Manage menu.

  3. Navigate to 192.168.49.1:8080/dash with a phone RC or 192.168.43.1:8080/dash with a Control Hub.

  4. Ensure that you have the Field view selected in the top right.

  5. You should see two lines and two circles being drawn: green for the target position and blue for your bot's actual position.

  6. Look for the SampleMecanumDrive in the right sidebar. Open that dropdown. You should be seeing two options: HEADING_PID and TRANSLATION_PID. Both options are located in the SampleMecanumDrive file.

  7. Open up HEADING_PID first. Just keep increasing kP until the robot starts to keep an accurate heading. This was around 8 in my experience, although your mileage may vary. You should not need to adjust kD and kI.

  8. Open up TRANSLATION_PID next. Once again, keep increasing kP until the robot starts to adjust itself and follows the path. This was also around 8 in my experience, although your mileage may vary. You should not need to adjust kD and kI.

  9. Once that's tuned, you should be done! Remember that any changes in Dashboard must be reflected in the appropriate file. So, you should copy-paste your numbers into the PID object in SampleMecanumDrive.java.

  10. Feel free to run the same tuning process with FollowerPIDTuner. This is encouraged for further accuracy.

  11. You should be done! Go on to the SplineTest to ensure that your following is accurate.

  12. Note: Earlier I mentioned that you should not need to use kI or kD for your follower PID's. Those are just basic guidelines. kD actually indirectly functions as a P Controller on pose velocity. Therefore, setting kD to a non-zero value (1 or something), IF you are using feedforward instead of velocity pid, may help keep your pose velocity constant across battery voltages. DO NOT set kD if you are using drive velocity PID as they will overlap and fight each other.

# Temporary Informal Video

Temporary informal video while the polished, edited videos are in production. Quality is atrocious but hopefully it helps 😃