摘要
使用前向神经网络 ,采用带阻尼的牛顿二阶学习方法 ,学习纯物质的饱和液体密度与温度的关系 ,在熔点到临界点的温度范围内 ,预测平均误差小于 0 .0 3%。适宜的网络工作区间 [amin,amax ]为 [0 .5 ,0 .7]
The feed forword neural network is used to study the relationship between the density of the pure saturated liquid matters and the temperature. The weighs of the neural network are updated by using the damped Newton second order method. The estimated average errors are less than 0 03% between the melting point and the critical point. The suitable working range [ a min ,a max ] is [0 5,0 7] for the network.
出处
《广西科学》
CAS
2000年第3期201-202,205,共3页
Guangxi Sciences
关键词
神经网络
预测
牛顿二阶学习
饱和液体密度
温度
neural network, density of liquid, estimation,Newton second order method
