A carregar...
Publicação
Procedural Generation of 2D Games
| Nome: | Descrição: | Tamanho: | Formato: | |
|---|---|---|---|---|
| 3.59 MB | Adobe PDF |
Orientador(es)
Resumo(s)
The main objective of this project is to develop a procedural generator of levels for 2D games, with the capacity of adapting the difficulty of the levels to a player’s skill in a specific game. Thus, in order to implement a procedural generator with the previously mentioned features, we intend to combine two techniques: procedural content generation and dynamic difficulty adjustment. Procedural content generation is a technique which has the purpose of creating content for a game. The game content generated can be anything related to the video-game in question (e.g. characters, items, terrain, levels). Dynamic difficulty adjustment is the name of the technique used to make adjustments to the game’s difficulty, depending on the overall progress of a player in a particular level. The procedural content generator developed uses the idea of rhythms of a level as its basis (Smith et al., 2009). This approach consists on describing a level as a sequence of actions that must be done to successfully conclude it. Our methodology differs from the classical rhythm-based approach, because instead of a sequence of single actions we rep resent a level as a sequence of classes of actions. A class of actions is a group of actions that have the same assumed difficulty, which is defined by a mechanic description (what keys to press to perform an action). For the generation of these sequences of classes of actions, it is used a genetic algorithm whose fitness function is able to evaluate the difficulty of a sequence, which allows it to generate rhythms for diverse levels with different difficulties. After the rhythm generation process, the resulting sequences of classes of actions are going to be passed as a parameters to a geometry generator, that is going to associate each of the class of actions to a level chunk, having, in the end, a new playable level (a group of level chunks). This approach was then tested with different games to demonstrate the generator’s capacity to generalize and, to prove our definitions of difficulty, we made some tests using search algorithms and human players to make this evaluation.
Descrição
Tese de mestrado, Engenharia Informática (Interação e Conhecimento) Universidade de Lisboa, Faculdade de Ciências, 2020
Palavras-chave
Geração procedimental de conteúdo Ajustamento dinâmico de dificuldade Jogos 2D Algoritmo genético Inteligência artificial Teses de mestrado - 2020