The Study of An Hybrid Learning Algorithm And An Hybrid Architecture Model In Agent-based Simulation

To improve learning efficiency, an hybrid learning algorithm is proposed to combine individual learning and group learning effectively; it improves agent´s ability and system´s intelligence level; to satisfy the requirements of complex adaptability, an hybrid architecture model is developed; in this model, a coordination control unit based on knowledge is proposed to coordinate the cognitive process and reactive process.