摘要
介绍了一种应用神经元网络建模仿真的系统参数优化设计方法 .采用正交试验设计法准备训练数据 ,采用 BP网络对试验数据建模 .利用网络模型仿真正交试验得到各参数对性能指标的影响曲线 ,通过直观分析可对系统参数组合进行优化设计 .将该方法用于某化工产品得到优化配方 。
A method for optimizing system parameters based on neural network modeling is propsed. A BP neural network is used as the model of system and is trained with orthogonal data set. Through intuitionistic analysis of the curves which are resulted from the model and showed the influences of parameters on the performances of system, the optimal combination of system parameters can be obtained. The method has been used to optimize the formulation of some chemical product, and proved to be highly effective.
出处
《信息与控制》
CSCD
北大核心
2001年第6期562-565,共4页
Information and Control
关键词
神经元网络
系统参数优化
数学模型
设计方法
optimization of parameter, neural network, system performance, orthogonal design
