In view of the limitation of practicality and applicability of generally used offline identification, we propose an active vibration control algorithm based on online control channel real-time identification by analyzing the basic control structure and identification strategy. We introduce random noise at the control output and achieve real-time online identification for the control channel using an FIR filter. A filtered-X algorithm is employed for the control. The controller architecture and algorithm process is illustrated. Experiments are performed on the active vibration control platform to assess the identification results. The results show that the proposed control algorithm is feasible and effective.