论文部分内容阅读
为了提高径向基函数(RBF)神经网络的泛化能力,提出了一种组合径向基函数神经网络,并通过对英文字母的识别进行了仿真验证。基于CMOS电路设计了该组合径向基函数神经网络,所有单元电路均采用HJTC 0.18μmCMOS数模混合工艺设计制造。通过PCB板实现了一个2×3的组合RBF神经网络,并对“一”和“1”的识别问题进行了验证。实验结果表明:该组合RBF神经网络电路结构简单,便于扩展和调节,提高了整个网络的泛化能力,为硬件实现更为复杂的组合径向基函数神经网络提供了可能。
In order to improve the generalization ability of Radial Basis Function (RBF) neural networks, a combined radial basis function neural network is proposed and verified by the recognition of English alphabets. Based on the CMOS circuit design of the combination of radial basis function neural network, all unit circuits are used HJTC 0.18μmCMOS digital-analog hybrid design and manufacture. A 2 × 3 combined RBF neural network is realized by PCB, and the identification of “” and “1 ” is validated. The experimental results show that the proposed RBF neural network is simple in structure, easy to extend and adjust, and improves the generalization ability of the whole network, which makes it possible to realize the more complex combined RBF neural network by hardware.