This project is an assignment for the Artificial Intelligence paper. For this assignment, I implemented a genetic algorithm to optimise the fitness of a population of vacuuming agents, a.k.a. cleaners, tasked with cleaning an area. Please see the engine.jpg for a demonstration.
The rules of the world are as follows: a cleaner can move forwards on backwards in discrete steps; it can rotate in either direction by 90 degrees; when driving over dirty area it automatically picks up the dirt (unless its bin is full); cleaners run on a battery that needs to be recharged periodically at a charging station, where also the bin is emptied. The objective is clean maximal area. Oh, and the final complication is that there is another population of cleaners on the field, competing for cleaning credits. My algorithm should find behaviours that lead to the most cleaning done by your cleaners. The engine is given by our lecturer Lech Szymanski. My jod is to build my_agent.py for using genetic algorithm to obtain a population of 40 cleaners with good performance.
For the detail of the algorithm and the evaluation, please see my_report.pdf.