基于神经网络的城市大气雾霾污染短时预测方法研究

Short-term Prediction Method of Urban Atmospheric Haze Pollution Based on Neural Network

城市大气雾霾污染短时预测性会影响人们日常出行,威胁人们生命健康,为了避免雾霾所造成的生命安全问题,提出基于神经网络的城市大气雾霾污染短时预测方法,该方法通过分析产生城市大气雾霾的原因,获取工业生产排放、煤炭排放和汽车尾气在合成雾霾时所需要的量,并将三者代入神经网络中,结合线性组合器构建雾霾短时预测模型,根据模型输出结果实现城市大气雾霾污染短时预测。实验结果表明,基于神经网络的城市大气雾霾污染短时预测方法应用后,短时预测准确性高。

Poor short-term prediction performance of urban atmospheric smog pollution can affect people s daily travel and threaten their lives and health.In order to avoid life safety issues caused by smog,a short-term prediction method of urban atmospheric smog pollution based on neural networks was proposed.This method analyzed the causes of urban atmospheric smog,and obtained the amount of industrial production emissions,coal emissions,and automobile exhaust required to synthesize smog.The three factors are substituted into a neural network and combined with a linear combiner to construct a short-term prediction model for smog pollution.According to the output results of the model,short-term prediction of urban atmospheric smog pollution is achieved.The experimental results showed that the short-term prediction accuracy of urban atmospheric haze pollution based on neural network is high after its application.