电流模反馈放大器参数分析神经网络模型

Artificial Neural-Network of Parameter Analysis of Current Model Feedbach Amplifie

文中从电路模拟的角度,探讨神经网络模型在电流模电路模拟中的应用。以影响电路性能的主极点的电容C_Z和反馈电阻R_f为样本,电路增益为输出样本,采用BP算法对多层感知器进行训练。在学习范围内,任意的给定负载电容Cz和反馈电阻R_f,在神经网络的输出端立即得到增益曲线,从而达到节省设计时间的目的。

In this paper, the applications of ANN (artificial neural-network) in the analog of current model circuits are in- vestigated. The load capacitor C_z and feedback resistance R_f of the main pole, which will affect the performance of the circuits, are chosen as samples. And the circuit's gain is chosen as output samples. The BP algorithm is the method to train multiplayer sensor. For any given load capacitor C_z and feedback resistance R_f, the gain curve at the output port of ANN is obtained to save time in the scope of the study.