Self-Driving Car Engineer Nanodegree Program

The goal of the project is run a car in the simulator with a Model Predictive Controller (MPC). The simulator provides way points for the track that should be followed by sending steering and acceleration commands to the simulator.

The video below shows a run with the simulator with the final model (click at the image to view the video -> redirect to youtube).

The figure below shows the state of the vehicle that is being used (screenshot from Udacity lesson):

• x and y of the vehicle
• psi: orientation of the vehicle
• v: speed

Next to that the following variables are also kept in the state:

• Cross Track Error (CTE)
• Error of the oriention (ePsi)

The figure below shows the equations that are being used in the model to calculate the next state (screenshot from Udacity lesson)

The MPC model being used is the one being discussed in the previous lesson of this course and exists of the following steps:

1. A set of waypoints is provided that should be followed
2. The current position and orientation of the car is provided
3. A vehicle model is provided that can predict the next state (position, orientation, speed, acceleration) of the vehicle
4. An optimizer function will find the best actuator values(throttle and steering value) to follow this path.

For the optimizer function, cost functions needs to be defined. In this model, the following costs functions have been used:

1. Cost functions for the cte en epsi errors
2. Cost function for the speed optimization
3. Cost functions to minimize the use of actuators
4. Cost function to minimize the gap between sequential steering angle (not the acceleration, it's a race car!!)
5. Cost function to relate the speed to the corner angle: the higher the angle, the lower the speed

T is the production horizon over which future predictions are made. T consists of:

• N the number of timesteps
• dt is the time between actuations

General guidelines are:

• T should be as large as possible, but not more than a few seconds because it does not make sense to predict too far in the future
• dt should be as small as possible
• The larger N, the more steps are involved which costs more computer power

Taken this into account the following have been tried

• N has been optimized between 5 and 25
• dt has been optimized betwen 0.05 and 0.5

For low Ns, the vehicle directly drove off the track, for higher N the vehicle never drove very stable. For lower dt there was not much improvement and higher dt caused too much ossilation.

The optimal values that have been found: N=10 and dt=0.15

The waypoints provided by the simulator are first converted to vehicle coordinates line 56-68(src/main.cpp)

A third order polynomial is than calculated to fit these waypoints (see helpers.h)

These polynomial is used to:

• Calculate the CTE and orientation errors
• In the cost optimization method to estimate the most optimal trajectory
• Calculate the waypoint for the simulator to show them (Yellow line)

To deal with the latency (100 ms) that occurs between the MPC controller and the simulator the following has been done:

The initial state of the optimizer function has been predicted after 100 ms. With this, the model was able to make robust prediction.

• cmake >= 3.5

• All OSes: click here for installation instructions

• make >= 4.1(mac, linux), 3.81(Windows)

  • Linux: make is installed by default on most Linux distros
  • Mac: install Xcode command line tools to get make
  • Windows: Click here for installation instructions

• gcc/g++ >= 5.4

  • Linux: gcc / g++ is installed by default on most Linux distros
  • Mac: same deal as make - [install Xcode command line tools]((https://developer.apple.com/xcode/features/)
  • Windows: recommend using MinGW

• uWebSockets

  • Run either install-mac.sh or install-ubuntu.sh.
  • If you install from source, checkout to commit e94b6e1, i.e.

    git clone https://github.com/uWebSockets/uWebSockets
    cd uWebSockets
    git checkout e94b6e1
    

    Some function signatures have changed in v0.14.x. See this PR for more details.

• Ipopt and CppAD: Please refer to this document for installation instructions.

• Eigen. This is already part of the repo so you shouldn't have to worry about it.

• Simulator. You can download these from the releases tab.

• Not a dependency but read the DATA.md for a description of the data sent back from the simulator.

1. Clone this repo.
2. Make a build directory: mkdir build && cd build
3. Compile: cmake .. && make
4. Run it: ./mpc.

The docker-compose can run the project into a container and exposes the port required by the simulator to run.

1. Clone this repo.
2. Build image: docker-compose build
3. Run Container: docker-compose up
4. On code changes repeat steps 2 and 3.

1. The MPC is recommended to be tested on examples to see if implementation behaves as desired. One possible example is the vehicle offset of a straight line (reference). If the MPC implementation is correct, it tracks the reference line after some timesteps(not too many).
2. The lake_track_waypoints.csv file has waypoints of the lake track. This could fit polynomials and points and see of how well your model tracks curve. NOTE: This file might be not completely in sync with the simulator so your solution should NOT depend on it.
3. For visualization this C++ matplotlib wrapper could be helpful.)
4. Tips for setting up your environment are available here
5. VM Latency: Some students have reported differences in behavior using VM's ostensibly a result of latency. Please let us know if issues arise as a result of a VM environment.

We have kept editor configuration files out of this repo to keep it as simple and environment agnostic as possible. However, we recommend using the following settings:

• indent using spaces
• set tab width to 2 spaces (keeps the matrices in source code aligned)

Please (do your best to) stick to Google's C++ style guide.

Note: regardless of the changes you make, your project must be buildable using cmake and make!

More information is only accessible by people who are already enrolled in Term 2 of CarND. If you are enrolled, see the project page for instructions and the project rubric.

• You don't have to follow this directory structure, but if you do, your work will span all of the .cpp files here. Keep an eye out for TODOs.

Help your fellow students!

We decided to create Makefiles with cmake to keep this project as platform agnostic as possible. We omitted IDE profiles to ensure students don't feel pressured to use one IDE or another.

However! I'd love to help people get up and running with their IDEs of choice. If you've created a profile for an IDE you think other students would appreciate, we'd love to have you add the requisite profile files and instructions to ide_profiles/. For example if you wanted to add a VS Code profile, you'd add:

• /ide_profiles/vscode/.vscode
• /ide_profiles/vscode/README.md

The README should explain what the profile does, how to take advantage of it, and how to install it.

Frankly, I've never been involved in a project with multiple IDE profiles before. I believe the best way to handle this would be to keep them out of the repo root to avoid clutter. Most profiles will include instructions to copy files to a new location to get picked up by the IDE, but that's just a guess.

One last note here: regardless of the IDE used, every submitted project must still be compilable with cmake and make./

A well written README file can enhance your project and portfolio and develop your abilities to create professional README files by completing this free course.