In this project, you will apply the skills you have acquired in this course to design a Proportional-Integral-Derivative (PID) controller to perform vehicle trajectory tracking. Given a trajectory as an array of locations, and a simulation environment, you will design and code a PID controller and test its efficiency on the CARLA simulator used in the industry.
The state code in this repository is aligned to run on the Udacity VM workspace. Refer to the classroom page Ubuntu VM Workspace - Overview to learn how to access the VM workspace and its restrictions and best practices.
However, to set up your local machine with the necessary tools, you must have either Windows Subsystem for Linux (WSL) or Ubuntu 20.04 or 18.04 LTS. Below is the list of tools installed in the Udacity VM workspace that you should install on your local machine.
-
CARLA simulator 0.9.9.4.
You can find more details at CARLA Quick Start Installation. The deb installation is the easiest way to get the latest release in Linux.sudo apt-key adv --keyserver keyserver.ubuntu.com --recv-keys 1AF1527DE64CB8D9 sudo add-apt-repository "deb [arch=amd64] http://dist.carla.org/carla $(lsb_release -sc) main" sudo apt-get update # Update the Debian package index sudo apt-get install carla-simulator=0.9.10-2
The installation directory must be /opt/carla-simulator/ on your Linux machine. To verify, open a terminal an launch CARLA as:
cd /opt/carla-simulator ./CarlaUE4.shThe Carla Simulator should launch in a few seconds. You can close it after verification.
-
NICE DCV Server.
This includes the Nvidia drivers along with CUDA libraries for the underlying Tesla T4 GPU.Sat Oct 14 15:31:45 2023 +---------------------------------------------------------------------------------------+ | NVIDIA-SMI 535.104.12 Driver Version: 535.104.12 CUDA Version: 12.2 | |-----------------------------------------+----------------------+----------------------+ | GPU Name Persistence-M | Bus-Id Disp.A | Volatile Uncorr. ECC | | Fan Temp Perf Pwr:Usage/Cap | Memory-Usage | GPU-Util Compute M. | | | | MIG M. | |=========================================+======================+======================| | 0 Tesla T4 On | 00000000:00:1E.0 Off | 0 | | N/A 31C P0 27W / 70W | 2093MiB / 15360MiB | 27% Default | | | | N/A | +-----------------------------------------+----------------------+----------------------+ +---------------------------------------------------------------------------------------+ | Processes: | | GPU GI CI PID Type Process name GPU Memory | | ID ID Usage | |=======================================================================================| | 0 N/A N/A 1055 G /usr/lib/xorg/Xorg 67MiB | | 0 N/A N/A 1521 G /usr/lib/xorg/Xorg 89MiB | | 0 N/A N/A 1669 G /usr/bin/gnome-shell 23MiB | | 0 N/A N/A 1948 C+G /usr/lib/x86_64-linux-gnu/dcv/dcvagent 398MiB | | 0 N/A N/A 3320 G ...sion,SpareRendererForSitePerProcess 30MiB | | 0 N/A N/A 4489 C+G ...aries/Linux/CarlaUE4-Linux-Shipping 1348MiB | +---------------------------------------------------------------------------------------+
dcv version # Output NICE DCV 2023.0 (r15487) Copyright (C) 2010-2023 NICE s.r.l. -
C++
gcc --version # Output gcc (Ubuntu 9.4.0-1ubuntu1~20.04.2) 9.4.0 -
Git
-
CMake and Make
-
Python3 and Pip v20.xx or above.
python3 --version # Output Python 3.8.10 -
ROS
-
Project specific dependencies
# Required for building project sudo apt-get install -y libgoogle-glog-dev libgtest-dev # Required for running project. # Install carla python package sudo python3 /usr/lib/python3/dist-packages/easy_install.py /opt/carla-simulator/PythonAPI/carla/dist/carla-0.9.9-py3.7-linux-x86_64.egg # Install python requirements pip install pandas matplotlib numpy pygame websocket-client
The sections ahead will guide you through the steps to build and run the project.
Open the VM workspace and log into the VM to practice the current project. Once you log into the VM, open a Terminal window.
Fork the repository to your Github account and clone the repository to the workspace using the commands below.
git clone https://github.com/udacity/nd013-c6-control-starter.gitChange to the project directory.
cd nd013-c6-control-starter/projectYou will find the following files in the project directory.
.
├── cserver_dir
├── install-ubuntu.sh
├── manual_control.py
├── pid_controller/ # TODO Files
├── plot_pid.py
├── run_main_pid.sh
├── simulatorAPI.py
├── steer_pid_data.txt
└── throttle_pid_data.txtStart the Carla server by executing the following shell script.
./run_carla.shThis file will, in turn, execute the /opt/carla-simulator/CarlaUE4.sh script.
Open another Terminal tab, and change to the nd013-c6-control-starter/project directory. Execute the following shell script to install the project-specific dependencies.
./install-ubuntu.shThis file will install utilities such as, libuv1-dev, libssl-dev, libz-dev, uWebSockets.
Change to the pid_controller/ directory.
cd pid_controller/Before you start coding, we strongly recommend you look at the rubric in your classroom, against which the human Mentor will review your submission. Your submission must satisfy all rubric criteria to pass the project; otherwise, the Mentor may ask you to re-submit.
Update the following files as per the classroom instructions. You will TODO markers as well in these files.
- pid_controller.h
- pid_controller.cpp
- main.cpp
Important: At this moment, it is important to save your work and push it back to the remote Github repository.
In the previous version of the project starter code, we had libcarla-install/ and rpclib/ directories inside the pid_controller/ directory. But, those directories are no longer needed in the current version of the starter code because the current CMakeLists.txt file has corresponding includes and libs added at /opt/carla-source.
To give some old context, when we had rpclib/ directory inside the starter files, we used to compile the rpclib library using the following commands.
cd pid_controller/
rm -rf rpclib
git clone https://github.com/rpclib/rpclib.gitThis library is a msgpack-rpc library written using modern C++. The goal of building this library was to provide a simple RPC solution. However, all of the above-mentioned steps are no longer needed in the current version of the project strarter code.
When you finish updating the project files, you can execute the project using the commands below.
# Build the project
# Run the following commands from the pid_controller/ directory
cmake .
# The command below compiles your c++ code. Run it after each time you edit the CPP or Header files
make# Run the project
cd ..
# Run the following commands from the nd013-c6-control-starter/project directory
./run_main_pid.shIf the execution fails silently, you can use ctrl + C to stop, and try again.
Another possible error you may get is bind failed. Error: Address already in use. In such a case, you can kill the process occupying the required port using the commands below.
ps -aux | grep carla
# Use the IDs displayed in the output of the last command.
kill id Re-check the rubric in the classroom and ensure that your submission satisfies all rubric criteria to pass the project. Once you are confident, submit the project.