Applying an Enhanced Path Finding Avatar for a Virtual Environment
Jui-Fa Chen, Wei-Chuan Lin, Lihao Yang · 2007
Abstract—There are several types of landform and objects in the virtual environment such as obstacles, move objects and items … etc. A path finding avatar should evaluate the whole environment properly, and design the policies or plans to find the path to the specific goal. With the goal of “moving to specific location”, this paper focuses on building the behavior model of a path finding avatar in a virtual environment. This paper proposes a hybrid method which combined the Potential Field Methods (PFM) and Virtual Force Field Methods (VFF) for a path planning avatar to evaluate the virtual environment. From the experimental results, the behavior avatar with the proposed method can move to the specific goal better than pure using the mechanism of PFM or VFF. In the future, the behavior models of original users could be kept in the behavior data-base for analyzing or training by this proposed method. In this way, this proposed method could apply the Bayesian Network mechanism for more advanced personalization. a self-learning mechanism. When the user leaves the virtual environment, the IBA replaces the original user and is in the active duration behavior. The IBA can utilize the learning experience that was collected in the passive duration behavior and behaved like the original user in the virtual environment. Index Terms—virtual environment, intelligent behavior avatar, potential field, virtual force field, path planning I.