将神经网络理论、预测理论及最优控制理论引入振动控制工程领域，提出了一种基于ＢＰ网络的主动控制方法．给出了该方法中神经网络的学习算法、模型辨识和预测公式及最优控制指标的选择原则．分析了它在建模和控制方面所具有的特点．经计算机仿真表明，该方法控制效果优于机械隔振和ＰＩＤ主动隔振，是行之有效的 The neural network theory, prediction theory and optimal control theory are introduced into the field of vibration control engineering, and a kind of active control method based on BP neural network is proposed. The learning algorithm of neural network, the model identification and prediction formula and the selection principle of the optimal control index are given. It analyzes the features it has in modeling and control. The computer simulation shows that this method is superior to mechanical vibration isolation and PID active vibration isolation, which is effective