臺灣博碩士論文加值系統

1998年暴雪娛樂製作發行的一款著名即時戰略遊戲，許多媒體給予了這款遊戲很高的評價，認為其是一部經典的即時戰略遊戲。遊戲提供了一個遊戲戰場，用以玩家之間進行對抗。這也是該遊戲以及所有即時戰略遊戲的核心內容。在這個遊戲戰場中，玩家可以操縱任何一個種族，在特定的地圖上採集資源，生產兵力，並摧毀對手的所有建築取得勝利。BWAPI(The Brood War Application Programming Interface)是一個免費且開放的 C++ 架構用來創造AI於Starcraft: Broodwar執行。在這篇論文中，我們運用類神經網路結合基因演算法來創造一個人工智慧，讓電腦能依照勝率來改變類神經的權重，進而讓他自己學習如何去贏得比賽。

StarCraft is a Real-Time War Strategy video game developed by Blizzard Entertainment in 1998. Real Time Strategy Games are one of the most popular game schemes in PC markets and offer a dynamic environment that involves several interacting agents. The core strategies that need to be developed in these games are unit micro management, building order, resource management, and the game main tactic. The player must reason about high-level strategy and planning while having effective tactics. Unfortunately, current games only use scripted and fixed behaviors for their artificial intelligence, and the player can easily learn the counter measures to defeat the AI. Enabling an artificial agent to deal with such a task entails breaking down the complexity of this environment. In this paper, we describe a system based on neural networks that controls what units should do in the game StarCraft. The system combined with genetic algorithm which can learn better way to play this game.

Chapter 1 7
Introduction 7
1.1 Motivation 8
1.2 About BWAPI 9
Chapter 2 10
Related Work 10
2.1 Neural Network 10
2.2 Genetic Algorithm 11
2.3 StarCraft 12
2.3.1 Protoss 13
2.3.1 Terran 13
2.3.2 Zerg 14
Chapter 3 15
Overview 15
3.1 System Architecture 15
3.2 Run the Game 17
Chapter 4 19
Methods 19
4.1 Neural Network Architecture 19
4.2 Input Value Normalization 21
4.3 Output 22
4.4 Genetic Algorithm 25
4.5 Tech/Build Tree 30
4.6 Build Place 31
Chapter 5 34
Experiment 34
5.1 Analysis of Original Type 34
5.2 Analysis of Two Independent Network 36
5.3 Analysis of Military Unit Strategy 38
5.4 Analysis of Adding Rules 40
5.5 Final Winning Ratio 41
Chapter 6 42
Conclusions 42
Bibliography 43

[1] J. Polvichai, M. Lewis, P. Scerri, and K. Sycara, Using Dynamic Neural Network to Model Team Performance for Coordination Algorithm Configuration and Reconfiguration of Large Multi-Agent Teams, C. H. Dagli, A. L. Buczak, D. L. Enke, M. Embrechts, and O. Ersoy, Eds. New York, NY: ASME Press, 2006.
[2] M. R. Khojasteh and A. Kazimi, “Agent Coordination and Disaster Prediction in Persia 2007, A RoboCup Rescue Simulation Team based on Learning Automata,” in Proceedings of the World Congress on Engineering 2010 Vol I., ser. WCE 2010. Lecture Notes in Engineering and Computer Science, 2010, pp. 122–127.
[3] I. Martinez, D. Ojeda, and E. Zamora, “Ambulance decision support using evolutionary reinforcement learning in robocup rescue simulation league,” in RoboCup 2006: Robot Soccer World Cup X, ser. Lecture Notes in Computer Science, G. Lakemeyer, E. Sklar, D. Sorrenti, and T. Takahashi, Eds. Springer Berlin / Heidelberg,
2007, vol. 4434, pp. 556–563.
[4] S. C. Bakkes, P. H. Spronck, and H. J. van den Herik, “Opponent modelling for case-based adaptive game AI,” Entertainment Computing, vol. 1, no. 1, pp. 27 – 37, 2009.
[5] Gabriel Synnaeve and Pierre Bessi` ere, “ A Bayesian Model for Plan Recognition
in RTS Games Applied to StarCraft,” in LPPA, Coll` ege de France, UMR7152 CNRS 11 place Marcelin Berthelot, 75231 Paris Cedex 05, France
[6] GitHub, https://github.com/bwapi/bwapi
[7] Bwapi An API for interacting with Starcraft: Broodwar (1.16.1) , https://code.google.com/p/bwapi/
[8] Wiki Artificial neural network, http://en.wikipedia.org/wiki/Artificial_neural_network

[9] Amirhosein Shantia, Eric Begue, and Marco Wiering, “Connectionist Reinforcement Learning for Intelligent Unit Micro Management in StarCraft,” Proceedings of International Joint Conference on Neural Networks, San Jose, California, USA, July 31 – August 5, 2011
[10] Aha, D. W.; Molineaux, M. and Ponsen, M. J. V. 2005, “Learning to win: Case-based plan selection in a real-time strategy game,” In ICCBR, 5–20.
[11] Albrecht, D. W, Zukerman, I. and Nicholson, A. E. 1998, “Bayesian models for keyhole plan recognition in an adventure game,” User Modeling and User-Adapted Interaction 8:5–47.
[12]Beal, M. J. 2003, “Variational algorithms for approximate Bayesian inference,” PhD. Thesis.
[13] Bessi` ere, P. Laugier, C. and Siegwart, R, “Probabilistic Reasoning and Decision Making in Sensory-Motor Systems,” Springer Publishing Company, Incorporated, 2008