[仿真]Neural-Based Dynamic Modeling of Nonlinear Microwave Circuits

Abstract—A neural network formulation for modeling nonlinear -RK- Fu<e  
microwave circuits is achieved in the most desirable ?5 [=(\/.  
format, i.e., continuous time-domain dynamic system format. vEz"xz1j!]  
The proposed dynamic neural network (DNN) model can be _2 osV[e  
developed directly from input–output data without having to i
MRwp+$  
rely on internal details of the circuit. An algorithm is developed `n?DU;,  
to train the model with time or frequency domain information. 1 I",L&S1  
Efficient representations of the model are proposed for convenient Ef13Q]9|  
incorporation of the DNN into high-level circuit simulation. U\!X,a*ts{  
Compared to existing neural-based methods, the DNN retains or CQDkFQq-dq  
enhances the neural modeling speed and accuracy capabilities, wJY'  
and provides additional flexibility in handling diverse needs of 57'4ljvYi  
nonlinear microwave simulation, e.g., time- and frequency-domain U_c
*6CK  
applications, single-tone and multitone simulations. Examples of gr-OHeid  
dynamic modeling of amplifiers, mixer, and their use in system t#eTV@-
 
simulati .. X+]G-  
i8HTzv"J  

未注册仅能浏览部分内容，查看全部内容及附件请先 登录 或 注册

ddddddddddddddddddddddddddddddddd

kkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkk

kkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkk