Path Planning with World Semantics
During my Final Year of University, I studied Games Artificial Intelligence,
for which my main project was to modify the standard A* algorithm to solve a game related artificial intelligence
problem. I chose to implement and use epsilon admissible A* to make computer controlled agents, or "bots" avoid
enemy held territory.
To demonstrate the algorithm, I wrote a small utility that allows you to "draw" your level
by placing walls and marking certain areas as difficult-to-pass which are dealt with by normal A*
in determining the best path between two points. Of course, you can also mark areas that are hostile; this
"cost" value is assessed during path calculation using the selected epsilon value. You can download the latest
version and also view some snippets from the project report.
The screen shots below show the same start and end points with the path calculated using
four different values for epsilon.
Primary Download Site:
To download it from this site: Click Here.
Secondary Download Site:
To download it from my (UK) ISP webspace: Click Here.
Path Planning
Home Page
Specification
 Design
and Implementation
Go to Top
|
