Conversion technology of attack-defense formations based on improved MOPSO algorithm

Effective conversion of attack-defense formations is one of the key technologies to enhance the level of robot scorer countermeasure. An improved algorithm of multi-objective particle swarm optimization (MOPSO) is proposed in this paper and applied to formation conversion system in SimuroSot. Taking the ball position as global extremum, this algorithm optimized the objective function with prediction and described goals movement according to the ball velocity. Through the continuous iterations, the optimal solution of velocity was presented for dynamic tracing so that formations can successfully convert. Experimental results show that the formation strategy can obviously reinforce the stability and cost less transitional time by use of improved MOPSO algorithm. In addition, the whole attack-defense ability of this team is enhanced.