Implement “Infastructure” the main{} that runs the “simulation” of the agents Agenti acting in the world:
State/world-situation representation
main(S(t), Ai(t))=S(t+1); main(S(t)){loop{for i : Ai(t)=action(Agenti,S(t));
S(t+1)= for i : apply action Ai(t) on S(t)}}
Implement agents that use the various Heuristic Search Algorithms...
Instantiate your infrastructure with a single agent to solve a puzzle of your choice (no TICTACTOE...)
Write a report describing your implementations and a statistics over sample runs .
Your statistic should compare the performance of different agents running different Search Algo and/or different heuristic evaluations
Instantiate your infrastructure with two agents to play a board game
(recommended chess)
(as before main(S(t), Ai(t))=S(t+1); but now agent1 does nothing when t=even and plays a move when t=odd, and viceversa for agent2.
Instructions and recommendations:
1.Keep your implementation as modular as possible, keeping various components (infrastructure, state representation and computation of agent actions effects, agents, search algo, Heuristic functions…) well separated from each other and/or encapsulated…
For example, it should be possible , given an implementation that plays chess, to use it to play a different game by simply changing the game model (representation and functions)
