摘要
为有效解决系统的最经济控制问题,本文提出将系统的经济收益问题转换为对系统控制结构和参数的优化问题。首先提出将网络代价的概念植入径向基函数神经网络(RBF网络)结构的优化中,对神经网络的隐层激活函数和隐层节点数进行选择;再用改进的遗传算法实现RBF网络参数的优化,从而实现神经网络的最经济控制;最后通过实例验证,表明设计的算法与BP网络的最经济控制相对比,具有明显的优越性。
In order to solve the problem of most economical control, this paper was proposed to transform economic efficiency of the system into optimization of the system＇s structure and parameters. Firstly the cost function of neural network was used into the structure optimization of RBF neural network, so that the hidden activation function of neuron the number of hidden layers were selected. Secondly improved GA was used into the parameters optinfization of RBF neural network .Lastly, an example was given to demonstrate the effectiveness of the programs.
出处
《微计算机信息》
2010年第2期37-38,43,共3页
Control & Automation
关键词
最经济控制
RBF神经网络
遗传算法
代价函数
Most Economical Control
RBF Neural Network
Genetic Algorithms
Cost Function
