论文部分内容阅读
BP神经网络通过调节连接权重可以实现以任意精度逼近非线性函数,利用这一特点可以对非线性函数关系进行拟合。偶氮苯聚合物的全开关特性曲线是非线性,很难用数学函数表达式来描述。因而本文首先介绍神经网络的基本原理和BP算法神经网络,然后BP神经网络应用于的全光开关特性曲线拟合,在MATLAB环境下,利用实验数据进行了实验测试,结果表明该方法处理数据精度高,拟合效果好。
The BP neural network can approximate the nonlinear function with arbitrary precision by adjusting the connection weight. This feature can be used to fit the nonlinear function. Azobenzene polymer full-switching characteristic curve is nonlinear, it is difficult to use mathematical expressions to describe. Therefore, this paper introduces the basic principles of neural networks and BP neural network algorithm, and then BP neural network used in the all-optical switch curve fitting, in the MATLAB environment, the experimental data using an experimental test results show that the method of dealing with data accuracy High, good fitting effect.