Aimed at the features that the emergent behavior is widely used and the occurred reasons are complex, a method of group emergent control based on data mining and fuzzy logic is proposed. Firstly, based on the traditional Swarm model, the dynamic data of agents under different parameters are obtained by traversing eight parameters influenced on the occurrence of emergent behavior. Meanwhile, linear regression and central clustering algorithm are used to recognize the type of emergent behavior under different parameters, and obtain the corresponding parameters of linear and circular emergent behavior. Then, the C4.5 algorithm is applied to precise parameters’ value range. Finally, the control of the occurrence of circular and linear multi-agent emergent behavior is realized by the fuzzy logic method. The effectiveness of the proposed control method is verified by simulation.