This paper develops a data-based hybrid driven control (DHDC) approach for a class of networked nonlinear systems compromising delays, packet dropouts and disturbances. First, the delays and/or packet dropouts are detected and updated online using a network problem detector. Second, a single-variable first-order proportional-integral (PI)-based adaptive grey model is designed to predict in a near future the network problems. Third, a hybrid driven scheme integrated a small adaptive buffer is used to allow the system to operate without any interrupt due to the large delays or packet dropouts. Forth, a prediction-based model-free adaptive controller is developed to compensate for the network problems. Effectiveness of the proposed approach is demonstrated through a case study.