Real time Strategy Games are one of the areas of AI that most present challenges related to planning and constraints satisfaction. In these games, the player needs to produce resources to develop his army in order to attack and defend against enemies. The choice of production goals has a direct impact in the strength of the army. In a previous work, we developed an approach based on Simulated Annealing that achieves quality goals by maximizing resource production. In this paper we extend our previous work by introducing a scheduling component. By intercalating planning and scheduling we were able to find better results than our previous approach. The experiments showed that our system was capable of competing with a human player without any restriction.
Naves, T. F., & Lopes, C. R. (2019). Maximization of the resource production in RTS games using planning and scheduling. In Proceedings of the 2015 International Conference on Artificial Intelligence, ICAI 2015 - WORLDCOMP 2015 (pp. 562–568). CSREA Press.