摘要
磁流变减摆器作为一种半主动控制元件被用于飞机前轮减摆控制,要想获得期望的控制性能,主要的困难就是建立准确的磁流变阻尼力计算模型和磁流变逆模型。本文在前面研究的基础上,基于自行设计的磁流变减摆器阻尼特性实验数据和修正Bouc-Wen模型。在此基础上,制造5000组随机电流和随机速度的作为输入,通过修正Bocu-Wen模型计算得到的阻尼力作为输出。将这三个变量的5000组数据,基于径向基神经网络,以前3000组作为训练数据,对径向基神经网络进行训练,以期获得能够逼近电流与速度、阻尼力之间的非线性函数关系。
As a semi-active control device,the magneto rheological(MR)damper is used to control the shimmy of aircraft nose wheels.In order to obtain the desired control performance,the main difficulty is to establish an accurate MR damping force calculation model and inverse model that can accurately input the current value.According to the previous research,based on the self-designed experimental data of damping characteristics of the MR damper and accurate modified Bouc-Wen model.On this basis,5000 groups of random current and random velocity were made as input,and the damping force calculated by modified Bouc-Wen model was used as output.The previous 3000 sets are used as training data to train RBF neural network,so as to obtain the nonlinear function relationship between current,velocity and damping force.
作者
林森
张继丰
马进
杜姗姗
Lin Sen;Zhang Jifeng;Ma Jin;Du Shanshan(Shenyang Customs Technology Center,Shenyang Liaoning,110000;Civil Aviation University of China,Shenyang Liaoning,110000)
出处
《电子测试》
2020年第17期51-52,73,共3页
Electronic Test
