Distributed Extended Kalman Filter (EKF) Simultaneous Localization and Mapping (SLAM) with known correspondence

Distributed Extended Kalman Filter Simultaneous Localization and Mapping with known data association.

The code in this repository was submitted in fulfillment of the requirements for the course '140472 - Distributed Systems for Measurement and Automation'.

Resources

• Probabilistic Robotics (Intelligent Robotics and Autonomous Agents). Sebastian Thrun, Wolfram Burgard, and Dieter Fox. 2005.
• Distributed Cooperative SLAM using Information Consensus Filter. Rajnikant Sharma, Clark N. Taylor, David W. Casbeer
• Decentralised SLAM with Low-Bandwidth Communication for Teams of Vehicles. Eric Nettleton et al.
• UTIAS Multi-Robot Cooperative Localization and Mapping Dataset

Repository Structure

• ./assets/ folder contains 9 datasets for the assignment
• ./include/ folder contains the implementation of the classes used for the simulation
• ./utils/ folder contains miscellaneous function for loading, resampling and animating the dataset

The root folder contains three main scripts: ./main.m, ./init.m and ./config.m

• ./config.m contains a Matlab structure with the user defined simulation parameters
• ./init.m loads the dataset, it resamples it and initializes the Agents in the simulation
• ./main.m is the file to execute to perform the simulation. It contains the main simulation loop and a call to the animation function

Results

The comparison of the Distributed EKF SLAM algorithm with the ground truth for the second dataset is displayed below:

Documentation:

Classes

• Actuator.m
• Agent.m
• ekfSLAM.m
• Message.m
• Sensor.m
• Server.m

Helper Functions

• cvt_rb_to_xy.m
• cvt_xy_to_rb.m
• find_transformation.m
• load_config.m