基于神经网络的主动隔振器

Active Isolator Based on Neural Networks

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。

An active control of vibration based on BP networks was proposed. This method synthesizes theories of neural networks, prediction control and optimum control into the field of vibration. The learning algorithm of networks, the formula of model identification and prediction, the rule of optimum control including their features were also introduced. The simulation proves its effect.