Simulated Robot Soccer Competition Policy Research And Realization

The multi-robots soccer competition, the typical question of the Distributed Artificial Intelligence research area, received more attention in recent years. The robot soccer involves many research areas, while simulation platform robot soccer research mainly concentrates on the dynamic and unforeseeable environment of the multi- robots soccer, single robot Agent decision-making and inference and overall team cooperation coordination.This article deals with the player strategy movement and field strategy of the simulation platform robot soccer system. The following are the main contents:(1) this article coordinates from the single player's movement and team's cooperation, has realized player's main strategy movement, sharpened player's technique and tactics ability.(2) this article provides the robot players with the limited definite movement set. According to the duty that system assigned, each player Agent chooses appropriate movement through decision system inference. By considering the player physical strength model, the article adjusts the player's role assignment strategy in order to ensure players' physical strength better restoration.(3) Utilizes the Particle Swarm Optimization multi-objectives optimization to deal with the run the position question while defending many player attacking, adapts the situation by the grain of subgroup algorithm randomness which changes unceasingly, strengthens destructive capability which attacks to opposite party.In the multi- robots soccer simulation platform environment, this article has arranged three experiments, carries on the examination separately to test the above contents, the statistical data demonstrates players' partial technique and tactics level, the physical strength arrangement and the defensive ability have improved.