Improve the Controller (e.g., PD Controller)

Improve the Controller

As mentioned in the previous chapter, instead of recreating the Python library with MATLAB script, we create a Python library that will run all necessary libraries needed to operate the JetRacer, then run it with MATLAB in the background.

Jetracer

• Requirements:

1. Jupyter libraries containing jetracer.nvidia_racecar must be working
2. Install pygame library on jetson nano with sudo pip install pygame
3. Create a python file that will import pygame, import NvidiaRacecar, read and write steering values.

Tabelle 5: Python Library Setup

#	Name	Description	Pictures
1	Main python libraries	These libraries are the root libraries of the JetRacer and must be available because other libraries will depend on them.	Fig. 6: Python libraries
2	Library for Control and steering	Since we can not use MATLAB directly on the Jetracer, we can, however, call a python code to run in the background on Matlab	Fig. 7: Controller library files
3	Python script	This script details how the inputs from the Gamepad is recorded and transmitted to Matlab thought steering_values.txt	Fig. 8: Controller library script

Matlab

After the above requirements are satisfied, we then proceed to call the library in the background as shown below:

Fig. 9: Matlab background library call

Results

zeigmess jtest etc