摘要
论文首先简述了主动磁轴承的工作原理及数学模型,比较了基于BP前馈网络和Elman递归网络的非线性辨识,并提出了一种加快神经网络收敛速度的辨识方案。通过对某单自由度磁轴承的仿真表明,该方案具有收敛速度快,模型精度高等优点。
The principle and mathematic model of active magnetic bearings（AMB） were discussed in brief firstly in this paper. Then a nonlinear identification method was compared based on BP feed forward network and Elman recurrent network, and a new identification method which can increase the convergence speed of the neural network was proposed. The simulation results of the AMB system showed that, the new method has the advantages include（such as） quicker convergence, higher accuracy and so on.
出处
《计算机与数字工程》
2013年第7期1119-1121,共3页
Computer & Digital Engineering
关键词
主动磁轴承
系统辨识
BP网络
ELMAN网络
active magnetic bearings, system identification, BP network, Elman network
