摘要
在雾霾天气准确预测的研究中,为更好的反映长春市空气质量状况,对长春市的PM2.5含量实测数据与同期的气象资料进行了分析。为科学准确地预测雾霾天气,根据BP人工神经网络可以逼近任意非线性函数的特点,提出了采用BP人工神经网络的雾霾天气预测系统。以长春市的PM2.5实时监测数据为时间序列数据样本,应用C++语言进行编程实现。最后建立了基于时间序列的BP神经网络雾霾天气预测模型。仿真结果表明,该模型能够较准确的预测雾霾天气。提高了预测的有效性和实用性。
In order to reflect the status of air quality in Changchun City, the actual data and meteorological data of PM2.5 levels during the same period are analyzed. To predict haze weather scientifically and accurately, according to the characteristics of BP artificial neural network, which can approximate any nonlinear function, a haze weather forecasting system is proposed based on the BP artificial neural network. Taking the real-time monitoring data of PM2.5 in Changchun City as the time series data sample, the system is realized by using C＋＋ language. Finally, a prediction model which is based on time series BP neural network to predict the haze weather is established. The simulation results show that the model can accurately predict the haze weather and improve the validity and practicability of the prediction.
出处
《计算机仿真》
CSCD
北大核心
2015年第1期402-405,415,共5页
Computer Simulation
