Abstract：Based on MATLAB toolbox,the analysis program of the BP neural network was built,and it was applied to evaluate PM2.5 and PM10. The three-layer BP neural network prediction system is constructed,and monthly data of two typical cities-Baoding and Lasa are used as training samples from October 2014 to January 2016. The prediction results of February and March in 2016 conform well to the published data. The study indicates that the analysis methods for PM2.5 and PM10 based on BP neural network are feasible and reasonable. And the results evaluation are objective and practical.
张达永,张祥,陈军. 基于MATLAB人工神经网络的典型大气PM2.5和PM10分析[J]. 气象水文海洋仪器, 2016, 33(3): 54-57.
Zhang Dayong,Zhang Xiang,Chen Jun. Analysis on typical PM2.5 and PM10 data based on MATLAB neural network. Meteorological Hydrological and Marine Instrument, 2016, 33(3): 54-57.
2016, Vol. 33 
