| 摘要: |
| 在空气污染大数据预处理的基础上，提出了一个基于深度学习的空气污染物浓度预测平台.该平台架构分为数据采集层、模型层和可视化界面层3个层次，分别实现了数据采集与处理，基于长短期记忆（LSTM）网络模型的污染物浓度预测，以及预测结果可视化的功能，通过对用户个性化模型参数的设置，实现不同时间段污染物浓度时间序列的预测. |
| 关键词: 深度学习 空气污染预测 长短期记忆(LSTM)网络模型 大数据 |
| DOI：10.3969/J.ISSN.1000-5137.2020.01.016 |
| 分类号:TP301 |
| 基金项目:国家自然科学基金（61802258）；上海市自然科学基金（18ZR1428300）；上海市科委创新项目（17070502800）；上海市教委项目（C160049） |
|
| An air pollutant concentration prediction platform based on deep learning |
|
LU Shuyi1, ZHANG Bo1, ZHANG Hanwen1, YU Hao1, GAO Haoran1, LIU Bo2
|
|
1.College of Information, Mechanical and Electrical Engineering, Shanghai Normal University, Shanghai 201418, China;2.Shanghai Super Computing Technology Co., Ltd., Shanghai 201203, China
|
| Abstract: |
| Based on the preprocessing of air pollution big data,a platform for predicting the concentration of air pollutant was proposed based on deep learning in this paper.The platform structure was divided into three levels,including the data acquisition layer,the model layer and the visual interface layer.What’s more,data collection and processing,pollutant concentration prediction based on long short-term memory (LSTM ) network model,and visualization of prediction results were realized by the three levels respectively.The platform could be customized by parameter configration to make time series prediction of pollutant concentration in different time periods. |
| Key words: deep learning air pollution prediction long short-term memory (LSTM) network model big data |
