[Inner States In Game Agent Systems]

Games C++ AI Date: Jun 2009

[BSc dissertation]

DOWNLOAD PAPER
The theoretical part of the project focuses on currently used techniques for obstacle avoidance, task-oriented behaviour and fuzzy logic in the field of games AI. A number of approaches are discussed and evaluated.


The artefact of the project is a strategy game Alien Farm, later developed into Stardust Colonies, where units have an inner state which reflects on their actions. This addresses a problem in most of today's strategies - units behave in the same way and display no reactions to how user plays a game.

Agents in the game use fuzzy logic to evaluate their current state and adjust their behaviour accordingly. The individualised behaviour was evaluated as an interesting feature by a number of players who tested the game.

Other aspects of games intelligence have also been experimented with - including obstacle avoidance in a continuous environment, finite-state machines, task representation and task-oriented behaviour.



{Please enable JavaScript in order to post comments}

Stardust Colonies

Stardust Colonies is a strategy game in which player-controlled units are aware of and affected by both their own memory and preferences, as well as by their colony's well-being.

A 2D iPhone Game Framework

The project reviews some of the existing design patterns and available algorithms used in iPhone games development. The application created is a template project for a 2D iPhone game.

Procedural Modelling

The paper provides an insight into procedural modelling techniques and uses. Procedural modelling is an alternative approach to modelling when high realism and fast real-time rendering of complex objects is desired.

pyCreeper

The main purpose of pyCreeper is to wrap tens of lines of python code, required to produce graphs that look good for a publication, into functions. It takes away your need to understand various quirks of matplotlib and gives you back ready-to-use and well-documented code.

Novelty detection with robots using the Grow-When-Required Neural Network

The Grow-When-Required Neural Network implementation in simulated robot experiments using the ARGoS robot simulator.

Fast Data Analysis Using C++ and Python

C++ code that processes data and makes it available to Python, significantly improving the execution speed.