Check out the new project site on Sourceforge: https://sourceforge.net/projects/easyagentsimulation/
(as of Nov. 15. 2014, this site is depricated.)
EAS is a Java-based simulation platform, developed as part of a research project at the Karlsruhe Institute of Technology. It comes with a broad range of pre-developed simulation scenarios, ranging from virtual agents acting in abstract environments to agents with a spatial representation in 2D or 3D. It is built upon a novel architecture using a separate "plugin layer" to communicate between the simulation engine and the simulated model. This allows to flexibly add or remove non-simulation-specific functionality such as visualization, chart drawing, trajectory painting etc. Studies show that the architecture is well-suited to guide beginners quickly to simulations of various levels of complexity. (The according research paper and documentation can be found under "Files"; Javadoc is available for most core classes and methods.)
Features
- Scheduler/Plugin Layer allowing to flexibly add non-simulation-specific functionality.
- 3D spatial/physical simulation.
- 2D grid/spatial/physical simulation.
- Abstract simulations.
- Generic sensors/actuators/agent brains.
- Artificial evolution.
- Machine learning.
- Finite state machines.
- Artificial neural networks.
- Complete sources and Javadoc