A Real-Time Simulation Mechanism for Unlimited Variability of Situations Based on Dynamic Event Coupling and Parameterized Animation against a Background Virtual World

Abstract

While the story plots in conventional RPG games 
are focused on its dramatic development, a system for education based on virtual world pursue diversity and unpredictability. The recent popularity of Open-World games such as Grand Auto Thief 5 over RPG games is due mainly to the fact their users tend to be immersed in the virtual world as its residents The monolithic authoring used in conventional narrative systems is not appropriate for edugames for language learning, 
our target application, which should provide the users with diverse experience. In an attempt to accomplish story plot variability and to solve this authoring scalability, we propose a situation simulation method based on dynamic coupling among independent events in contrast to monolithically pre-authored situations. An overarching event corresponding to an agent’s goal is decomposed and meaningfully connected into a multi-event plan in a search-based planning, which tend to be 
coincidentally coupled with other agents’ plans in their 
execution against the background world. This simulation is implemented with multiple real-time priority queues corresponding to agents’ plans. The actions constituting those queues each are assembled in terms of flexible parameterized motions which allow many actions to be performed in parallel. To reflect the potential discrepancy between the background world conditions and their corresponding knowledges of the 
agents, our simulation conducts the validity testing and 
considers those agents’ recognition on the preconditions of events in execution. We implement several complex multi-event situations to demonstrate the variability and scalability of situations resulting from our simulation mechanism based on the dynamic event coupling via the background world.