基于自编码网络的空气污染物浓度预测

An Air Pollutant Prediction Model Based on Auto-Encoder Network

深度学习为城市空气污染物浓度预测提供了更为强大的数据拟合能力,为空气污染预测提供全新的智能计算方法.为此,提出了一个基于自编码神经网络的污染物浓度预测模型AEPP(auto-encoder-based pollutant prediction).该模型包括编码器和解码器两个部分.其中,编码器用于提取出时间序列污染物浓度数据分布特征,即语境向量;解码器利用提取的特征预测未知时间内污染物浓度数据.模型中编码器和解码器采用多层LSTM(long short-term memory)模型结构,实现长时间依赖预测目标.实验表明,提出的模型可以提高对污染物浓度的预测水平.

In this paper, an autoencoder-based pollutant prediction (AEPP) model is proposed based on the auto-encoder neural network, which is composed of an encoder and a decoder. First, the encoder extracts the distribution characteristics of the time series of pollutant concentration data, namely the context vector. Secondly, the decoder uses the extracted characteristics to predict the pollutant concentration data in the next unknown time. Both the encoder and the decoder in the model can adopt several LSTM structures for long-time prediction. Experiments show that the AEPP model proposed in this paper can improve the effect of pollutant prediction.