| BP神经网络的硬件实现 |
| |
| 引用本文： | 任鲁涌,庞维珍.BP神经网络的硬件实现[J].山东理工大学学报,1996(4). |
| |
| 作者姓名： | 任鲁涌 庞维珍 |
| |
| 作者单位： | 山东工程学院计算机工程系(任鲁涌)，天津大学(庞维珍) |
| |
| 摘 要： | 本文提出了一种采用MOS晶体管的电流模式连续时间模拟BP（Backward Propagation）神经网络的硬件实现新方法。硬件系统适合利用VLSI技术,具有片内误差反向传播学习功能,能简化BP新算法修正权值,对所设计的电路提供了PSPICE仿真结果,结果表明硬件实现方法的可行性。
|
| 关 键 词： | 神经网络 BP算法 突触电路 细胞体电路 |
Hardware Realization of Backward Propagation (BP) Neural Network |
| |
| Ren Luyong,Pong Weizhen.Hardware Realization of Backward Propagation (BP) Neural Network[J].Journal of Shandong University of Technology:Science and Technology,1996(4). |
| |
| Authors: | Ren Luyong Pong Weizhen |
| |
| Institution: | Ren Luyong Pong Weizhen( Department of Computer Engineering ) ( Tianjin University ) |
| |
| Abstract: | This paper proposed a new method for hardware realization of BP neural network. This method, using current-mode continuously time-modeling, was well adaptive to the VLSI technique, had the in-chip error backward propagation learning function and simplified the corrected weight value of BP new computation, so that provided the circuits designed by the authors with PSPICE emulational result. |
| |
| Keywords: | Neural network BP computing method Sudden-touch circuit Cell-body circuit |
| 本文献已被 CNKI 等数据库收录！ |
