基于小波神经网络的地铁车站内CO_2浓度预测控制  

CO_2 Density Predication and Control of Subway Station Based on Wavelet Neural Network

作　　者：王长涛[1] 王德宝[2] 

机构地区：[1]沈阳建筑大学信息与控制工程学院,辽宁沈阳110168 [2]沈阳军区空军工程质量监督站,辽宁沈阳110015

出　　处：《科技广场》2011年第9期15-18,共4页Science Mosaic

摘　　要：文章提出了将小波函数引入神经网络预测模型,对一个具有非线性、时变性、大滞后性特点的地铁站台内的CO2含量进行预测控制。而小波神经网络系统结合了小波分析和传统神经网络的优点,且可不断吸收环境新信息,有良好的函数学习能力和推广能力。最后实现了对具有大干扰性、大滞后性和不确定随机干扰因素的CO2含量进行了精确预测。Wavelet function is in traduced into neural network prediction model is proposed to predict the carbon dioxide content of the subway platform as nonlinearity,time varying volatility and strong time delay.The Wavelet neural network system abstracts the advantages of wavelet analysis and traditional neural network,and has the capability of function learning and promoting for continuously absorbing environmental information as well.In this way,the carbon dioxide content which is disturbed by strong interference,time delay and uncertainty can be predicted accurately.

关 键 词：小波神经网络 地铁站台 CO2含量 预测 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]