Objective
This assignment focuses on implementing and analyzing an Extended Kalman Filter (EKF) for fusing IMU and GPS data to improve the localization accuracy of TurtleBot3 in a simulated environment. You will create a sensor fusion node, visualize the results, and experiment with different noise covariance matrices to understand their impact on the robot’s state estimation.
Task: IMU and GPS Sensor Fusion for TurtleBot3
Task Details
- Set Up the Sensor Fusion Node:
- Utilize the provided EKF implementation to fuse data from TurtleBot3’s IMU and GPS sensors. But you might be needing to upgrade it.
- Ensure that the EKF node processes the sensor data and outputs an accurate estimation of the robot’s position and orientation.
- Create a Custom Launch File:
- Develop a ROS 2 launch file to start the TurtleBot3 simulation along with the EKF sensor fusion node.
- The launch file should be configured to include all necessary parameters and topics for IMU and GPS data inputs.
- Visualize the Fused Data:
- Use RViz to visualize the robot’s estimated position and orientation as calculated by the EKF.
- Display the raw IMU and GPS data alongside the fused output to demonstrate the improvement in localization accuracy.
- Experiment with parameters of EKF
- Test the EKF with three different sets of process noise covariance and measurement noise covariance matrices.
- Document the behavior of the robot’s state estimation under each set of values, focusing on how the changes in parameters affect the accuracy and stability of the EKF.
- Analyze and Document the Results:
- Provide a detailed analysis of the impact of different Q and R values on the EKF’s performance.
- Include screenshots or recordings from RViz showing the robot’s path and the fused sensor data for each set of parameters.
Submission Process
- Create Files:
- Navigate to the module_6_assignment package.
- Create the required files for the EKF node, the custom launch file, and the documentation.
- Document Your Work:
- Create a README.md file in the module_6_assignment package.
- Provide details about the files you created, including explanations of the code and the commands needed to run your sensor fusion node and visualizations.
- Submit Your Assignment:
- Push your changes to your forked repository.
- Provide your repository link in the assignment submission text area.
- Note: Ensure you press the “Start Assignment” button when you see the page (as it takes time to generate the pages).
- Wait for Review:
- Wait for the instructors to review your submission.
Learning Outcome
By completing this assignment, you will:
- Understand the principles of sensor fusion using an Extended Kalman Filter (EKF).
- Gain hands-on experience with fusing IMU and GPS data to improve robot localization.
- Learn how to configure and tune an EKF for optimal performance in a simulated environment.
- Develop the skills to visualize and analyze fused sensor data using ROS 2 and RViz.