长短时神经网络在空气污染预报中的研究进展

Research Progress of Long and Short-Term Neural Networks in Air Pollution Forecasting

摘要: 空气污染预报是一项复杂的系统工程，是当今环境科学研究的热点，国内外已有将长短时神经网络(LSTM)用于不同项目预测的研究。是20世纪80年代以来人工智能领域兴起的研究热点。它在神经网络的基础上升级，延长了数据范围，具有时序性。本论文旨在综述长短时神经网络(LSTM)在空气污染预报中的研究进展，探讨其在空气污染预报领域的未来展望。

Abstract: Air pollution forecasting is a complex system engineering that is a hot spot in environmental science research, and there have been studies on the use of long and short neural networks (LSTM) for the prediction of different projects at home and abroad. It is a research hotspot in the field of artificial intelligence that has emerged since the 80s of the 20th century. It is an upgrade on the basis of a neural network, which extends the data range and is temporal. The purpose of this paper is to review the research progress of LSTM in air pollution forecasting and discuss its future prospects in the field of air pollution forecasting.