Research on Game Virtual Character Behavior Decision Driven by Artificial Intelligence Algorithms

Abstract

Online games have achieved sustained and rapid growth in various aspects such as game variety, quantity, genre, market size, and number of practitioners. The significance of behavioral Decision model is to make virtual roles have the ability to be as similar as possible to the behavior of real entity objects. Therefore, this article attempts to conduct research on AI (Artificial Intelligence) algorithmic driven game virtual character behavior decision-making. Research has shown that the algorithm proposed in this paper can train states in a stable state space faster in behavioral decision-making experiments. In the early learning stage, the advantages of the algorithm are not obvious. With the progress of learning, the Rate of convergence of the algorithm will gradually accelerate. On the whole, the Rate of convergence of the algorithm in this paper is higher than that of ACA (Ant Colony Algorithm) algorithm. This article takes virtual football competition as an example to construct a virtual football competition scene and conduct research on the behavior decision-making module of virtual players participating in football competition. The aim is to make virtual players have a certain degree of autonomy and intelligence, and their behavior decision-making module has a good sense of reality.