( l ) in order to conquer the complexity of controller , the direct adaptive neural network control is induced into active vibration control system and the digital simulation is made on the two - stage isolation mounting . the error channel is replaced with sgn function in this method . the results indicate the theoretical feasibility of the method ( 1 )将直接自适应神经网络控制方法引入双层隔振振动主动控制系统,该方法将误差通道纳入神经网络用sgn函数替代,不需要预先或在线辨识系统的数学模型,从而降低了控制系统的设计复杂性,在双层隔振台架上进行了数字仿真研究,结果表明了该方法的理论可行性。
Deducing the transfer function of the system , one controller is constructed and it has the wide frequencies isolation performance and zero force transfer rate . the active vibration isolation simulation is made and the results show that the two - stage isolation mounting with the controller constructed using the method mentioned in this paper has perfect isolation property and this provide new way for controller design . 4 通过对双层隔振振动主动控制系统模型的力传递函数的深入研究,构造了一种具有宽带隔振性能的理论上具有零力传递率的控制器,并进行了数字仿真研究,取得了理想的仿真效果,为振动主动控制控制器的设计提供了新的思路。
The experiments of active vibration control are made using direct adaptive neural network controller and direct adaptive pd neural network controller on the two - stage isolation mounting , in addition , the experiment of noncollocated vibration control is made using the two mentioned network controller . all experiment results indicate the effectivity of adopted method 在双层隔振模拟实验台架上进行了直接自适应神经网络振动主动控制实验和直接自适应pd神经网络振动主动控制实验,另外进行了上述两种神经网络的非并置振动主动控制的实验研究,实验结果验证了所采用方法的有效性。