This project contains the simulator of the 2016 OSU Robotics REU multiagent coordination team. The initial implementation will be a more basic gridworld that will be used for theoretical research. These models will then be translated to a physical platform, the Pioneer robots. There is the possibility for further work on more realistic simulator once the basic gridworld is complete.
- Connor Yates, firstname.lastname@example.org
- Abby Van Soest, email@example.com
- Tobi Ogunyale, firstname.lastname@example.org
This project requires FANN to be installed on the system to implement the neural network to control the agent policy.
Building and Running
The provided makefile has been tested so far only on Linux systems.
To run a program that relies on FANN, it is necessary to set the proper environment variables for the runtime linker.
FANN installs its libraries in
/usr/local/lib by default, so if using the default system installation, the runtime linker may not find the necessary libraries.
This can be fixed temporarily in a terminal with
or my adding the FANN library path to the ld config directory.
This is traditionally in
/etc/ld.so.conf.d/, and requires a root user invocation of
ldconfig after adding the config file.