Unscented Kalman Filter Project

Self-Driving Car Engineer Nanodegree Program

In this project utilize an Unscented Kalman Filter to estimate the state of a moving object of interest with noisy lidar and radar measurements. Passing the project requires obtaining RMSE values that are lower that the tolerance outlined in the project rubric.

This project involves the Term 2 Simulator which can be downloaded here

This repository includes two files that can be used to set up and intall uWebSocketIO for either Linux or Mac systems. For windows you can use either Docker, VMware, or even Windows 10 Bash on Ubuntu to install uWebSocketIO. Please see this concept in the classroom for the required version and installation scripts.

Once the install for uWebSocketIO is complete, the main program can be built and ran by doing the following from the project top directory.

1. mkdir build
2. cd build
3. cmake ..
4. make
5. ./UnscentedKF

Tips for setting up your environment can be found here

Note that the programs that need to be written to accomplish the project are src/ukf.cpp, src/ukf.h, tools.cpp, and tools.h

The program main.cpp has already been filled out, but feel free to modify it.

Here is the main protcol that main.cpp uses for uWebSocketIO in communicating with the simulator.

INPUT: values provided by the simulator to the c++ program

["sensor_measurement"] => the measurment that the simulator observed (either lidar or radar)

OUTPUT: values provided by the c++ program to the simulator

["estimate_x"] <= kalman filter estimated position x ["estimate_y"] <= kalman filter estimated position y ["rmse_x"] ["rmse_y"] ["rmse_vx"] ["rmse_vy"]

---

Thoughts on the Project

When computing mean and covariance matrix from sigma points for radar, computing them before or after mapping the prediction to the measurement space does make a difference. The mapping corresponds a nonlinear transformation. The mean and covariance matrix before the mapping are different from the mean and covariance matrix after the mapping. Compute them both. You'll need them in the final update step.

Bugs Encountered

Caution needs to be taken in multiple places for normalizing angles, otherwise nan or inf will be introduced to the simulation results, causing program to hang.

RMSE accuracy

All hyperparameters are relevant to the final RMSE estimation. Fine tuning them is necessary.

Remember that Kalman Filter is based on the linearity assumption. The only nonlinearity is introduced in radar update step. But in order to use sigma point method, we have to use sigma points method in predict step. It's not that we are changing the assumption of linearity, just to accommodate the sigma points method.

So for Lidar update which does not introduce nonlinearity we can use linear transformation as usual for predict step. Thus two versions of prediction function can be implemented here. But in my code I have not yet done that.

Using linear model in Lidar update improves the RMSE accuracy for dataset 1.

Other Important Dependencies

• cmake >= 3.5
  • All OSes: click here for installation instructions
• make >= 4.1 (Linux, Mac), 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
  • Windows: recommend using MinGW

Basic Build Instructions

1. Clone this repo.
2. Make a build directory: mkdir build && cd build
3. Compile: cmake .. && make
4. Run it: ./UnscentedKF Previous versions use i/o from text files. The current state uses i/o from the simulator.

Editor Settings

We've purposefully kept editor configuration files out of this repo in order 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)

Code Style

Please stick to Google's C++ style guide as much as possible.

Generating Additional Data

This is optional!

If you'd like to generate your own radar and lidar data, see the utilities repo for Matlab scripts that can generate additional data.

Project Instructions and Rubric

This 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.

How to write a README

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