论文部分内容阅读
针对前向正则模糊神经网络引进K-拟可加积分和K-积分模概念,应用积分转换定理研究了该网络在K-积分模意义下对模糊值简单函数类的泛逼近能力,进而在有限K-拟可加测度空间上,借助模糊值简单函数为桥梁获得了前向正则模糊神经网络依K-积分模对(?)-可积有界模糊值函数类仍具有泛逼近性.该结果表明前向正则模糊神经网络对连续模糊系统的逼近能力可以推广为对一般可积系统的逼近能力.
In this paper, we introduce the concepts of K-quasi-additive-addable integrals and K-integralsignals for forward-oriented regular fuzzy neural networks. By using the integral transformation theorem, we study the general approximation ability of this network to fuzzy function simple functions in the sense of K-integrals. K-Quasi-additive-measurable space, the forward regular fuzzy neural network is obtained by the fuzzy simple value function for the bridge, which still has universal approximation by the K-integral modulo (-) - integrable bounded fuzzy valued function class. It is shown that the approximation ability of the forward regular fuzzy neural network to continuous fuzzy systems can be generalized to the approximation capability of the general integrable system.