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

Abstract—A neural network formulation for modeling nonlinear N;`[R>Z~  
microwave circuits is achieved in the most desirable mTuB
*  
format, i.e., continuous time-domain dynamic system format. c!u}KVH  
The proposed dynamic neural network (DNN) model can be |n6Q  
developed directly from input–output data without having to PqJ*   
rely on internal details of the circuit. An algorithm is developed 3!oQmG_T  
to train the model with time or frequency domain information. :rs\ydDUF  
Efficient representations of the model are proposed for convenient <%3SI.  
incorporation of the DNN into high-level circuit simulation. 1={Tcq\]  
Compared to existing neural-based methods, the DNN retains or | 8L`osg  
enhances the neural modeling speed and accuracy capabilities, }jY[| >z  
and provides additional flexibility in handling diverse needs of +S5_J&~  
nonlinear microwave simulation, e.g., time- and frequency-domain #L IsL  
applications, single-tone and multitone simulations. Examples of )x:j5{>(  
dynamic modeling of amplifiers, mixer, and their use in system 4wj|  
simulati .. (8-lDoW  
"O*W]e  

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