欢迎您！东篱公司
退出
-
申报数据库
-
立项数据库
-
成果数据库
-
项目获奖数据库
中文摘要：
研究了一种非线性系统分析的神经网络算法,提出并证明了该算法的收敛性定理,为学习率的取值范围提供了理论依据.解决了BP算法存在局部极小的问题,并给出了该算法的应用实例.研究结果表明,对于随机给定的初始点,该算法都能稳定收敛到它的一个实根,计算精度可控,而且能得到高精度解,因此,该算法是有效的.此外算法还可以用来解多元非线性方程和线性方程组.
英文摘要：
The neural-network arithmetic to solve nonlinear systems was traversed. The convergence theorem of the arithmetic was presented and proved. This theorem gives theory gist to learning rate range. The local teeny problem of the BP arithmetic is overcome. The application examples are provided. The results show that the neural network can converge to a real solution stably for an arbitrarily given initial point of nonlinear systems of equations. The precisi.on can be controlled and the high precision, solution can be obtained. Therefore, this neural network arithmetic is effective. Furthermore, the arithmetic can also solve nonlinear equations in many variables and linear systems of equations.
同期刊论文项目
同项目期刊论文
期刊信息
位置：