Pareto frontier optimization in soccer simulation using normalized normal constraint

Abstract

Attacking is one of the popular tactics usually chosen by a soccer coach. With attacking effectively, chances to score goals can be enhanced as many as possible. Better attack needs good ball passing, and that is the purpose and focus of this research. Previous robotic soccer simulations employed simple weighting method with simple criteria and does not produce optimal ball passing. To overcome this problem this research proposes a set of criteria representing a more realistic situation. This set of criteria is formulated in and appropriate objective function which is optimized using pareto frontier and Normalized Normal Constraint. From the result of this experiment, it can be seen that this proposed method has a realibility of 20% higher that that of the previous method. And it has a better passing success rate with a 75% ball possession. Rather than that of the previous method with a 25% ball possession.