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

Abstract—A neural network formulation for modeling nonlinear ?0npEz|  
microwave circuits is achieved in the most desirable A1<k1[5fJ  
format, i.e., continuous time-domain dynamic system format. j<$R4A1  
The proposed dynamic neural network (DNN) model can be +(AwSh!  
developed directly from input–output data without having to 3LrsWAz'  
rely on internal details of the circuit. An algorithm is developed 2k3 z'RLG  
to train the model with time or frequency domain information. lS3 _Ild  
Efficient representations of the model are proposed for convenient EVSK
8T,  
incorporation of the DNN into high-level circuit simulation. oY
qE*mA  
Compared to existing neural-based methods, the DNN retains or fm6]CU1^  
enhances the neural modeling speed and accuracy capabilities, :
bw6k  
and provides additional flexibility in handling diverse needs of s?<!&Y  
nonlinear microwave simulation, e.g., time- and frequency-domain 0=,'{Vz}A  
applications, single-tone and multitone simulations. Examples of kh&_#,  
dynamic modeling of amplifiers, mixer, and their use in system m~Me^yt>}  
simulati .. .jum "va%  
QX/]gX  

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