A set of improved equations for dynamic feedforward neural networks is introduced to represent the behavior of the integrated circuit (IC). The rationality of the improved equations is elucidated by studying the relationship between the circuits and the equation parameters. In order to apply the dynamic feedforward neural networks to the design of intelligent IC, the method of modeling and simulation with VHSIC hardware description language is presented, as an important step for hardware implementation. The simulated results indicate that the new dynamic feedforward neural networks can better approximate to the behavior of a dynamic electrical system.
Citation:
Xi Chen, Gaofeng Wang, Wei Zhou, Jiangfeng Xu, Qinglin Zhang, "Modeling Dynamic Feedforward Neural Networks with VHDL*," imsccs, vol. 1, pp.785-788, 2006 First International Multi-Symposiums on Computer and Computational Sciences, 2006