A neural network based active control system was proposed to reduce the low frequency noise radiation of the simply supported plate. Feedback control systems with an identifier and without an identifier were built, respectively, in which neural network controller (NNC) and neural network identifier (NNI) were applied. Multi-frequency control in frequncy domain was achieved by simulation through the neural network based control systems. Both control systems have their advantages, and the former has high precision while the latter is easy to operate. The simply supported plate active vibration control experiment system was constructed. The control system without an identifer was applied to the experiment. The neural network based control algorithm was shown to perform effectively. The experiment lay a solid foundation of the engineering application of sturcture active vibration control.