摘要
研究了BP神经网络用于大气中可吸入颗粒物(PM_(10))及SO_2、NO_2预测的可行性。针对1999年至2001年的杭州市区大气中SO_2、NO_2和PM_(10)的实测数据,通过选用合适的BP网络,进行训练,得到预测模型,训练结果及结果检验表明BP网络适用于大气中SO_2、NO_2和PM_(10)的预测,并且该预测模型具有良好的适应性。
The application of BP neural network to predict PM_~10 , SO_2 and NO_2 pollutions is researched. After the design, the initialization and the training of BP neural network, the system of prediction is developed. The examples to predict the pollutions of PM_~10 , SO_2 and NO_2 in the Hangzhou city zone show that the BP neural network can be used to predict the PM_(10) , SO_2 and NO_2 pollutions if an appropriate network is constructed.
出处
《仪器仪表学报》
EI
CAS
CSCD
北大核心
2004年第z1期780-781,788,共3页
Chinese Journal of Scientific Instrument
