An intelligent model of the electromagnet-drived active mass damper(EMD) system using good nonlinearity processing of fuzzy neural network(FNN) is proposed. The EMD system is an innovative structural vibration control system in recent years. When the input voltage of the EMD system is beyond some frequency and extent, a series of performance tests verified that the experimental results didn’t agree well with the predications based on the current theoretical model of the force-electricity relationship of the EMD system. Considering the driving principles of EMD system, the active control force of the system is mainly relevant to the voltage of the electromagnetic coil and velocity of the EMD equipment. So, the FNN is utilized to identify the relationship between voltage, velocity and active control force of the EMD system in the paper. Firstly, a group of test results are taken as the training data of the FNN to model the dynamic behavior of a bench scale EMD. Then, another group of test results are taken to check the effect of the proposed model. The simulation results indicate that the predications based on the proposed intelligent model of the EMD system agree well with the experimental results under the generalized condition, and the intelligent model has better performance than the current model of force-electricity relationship.