论文部分内容阅读
利用实验数据，建立了特定转速下弹性连杆机构振动主动控制系统的动态前向神经网络模型．辨识过程分两步进行，首先实现串并联辨识，在串并联辨识成功的基础上实现并联辨识．系统辨识的结果表明，这种动态前向神经网络模型明显优于传统的ＫＥＤ理论分析模型．
Based on the experimental data, a dynamic feedforward neural network model of the active vibration control system of an elastic link mechanism under a specific speed is established. The identification process is carried out in two steps. Firstly, serial-parallel identification is realized, parallel identification is realized on the basis of the success of serial-parallel identification. The results of system identification show that this dynamic feedforward neural network model is obviously superior to the traditional KED model.