基于RBF神经网络的SO_2的预测  

Prediction of Sulfur Dioxide in Atmosphere Based on RBF Neural Network

作　　者：吕元锋[1] 田丽[1] 杭孟荀[1] 周峰[1] 黄世伟[1] 

机构地区：[1]安徽工程科技学院,安徽芜湖241000

出　　处：《兵工自动化》2005年第5期67-68,共2页Ordnance Industry Automation

摘　　要：基于RBF神经网络的SO2的浓度预测,选取一定的历史数据建立径向基函数神经网络训练模型,进行拟合训练。将芜湖市1993年到2001年大气SO2的浓度历史数据用于径向基函数神经网络,建立训练网络模型,通过训练优化提高训练可靠性。再用该模型对芜湖市大气中SO2的浓度进行预测。For the prediction of SO2 density based on the radial base function （RBF） neural network, according to the historical data of SO2 density, RBF neural network model is established to fit training. The historical data of SO2 density in atmosphere from 1993 to 2001 of Wuhu city was applied to RBF neural network, the network model for training was set up, and the model was trained and optimized to improve the dependability of the training. The density of SO2 in atmosphere of Wuhu city was predicted with the model.

关 键 词：径向基函数 神经网络 二氧化硫 浓度预测 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程] X510.2[自动化与计算机技术—控制科学与工程]