基于预测数据特征的空气质量预测方法

doi:10.6040/j.issn.1672-3961.0.2019.404

山东大学学报 (工学版) ›› 2020, Vol. 50 ›› Issue (2): 91-99.doi: 10.6040/j.issn.1672-3961.0.2019.404

基于预测数据特征的空气质量预测方法

高铭壑¹(),张莹^1,*(),张蓉蓉¹,黄子豪¹,黄琳焱¹,李繁菀¹,张昕²,王彦浩¹

1. 华北电力大学控制与计算机工程学院, 北京 102206
2. 长春理工大学计算机科学技术学院, 吉林长春 130022

收稿日期:2019-07-18 出版日期:2020-04-20 发布日期:2020-04-16
通讯作者: 张莹 E-mail:1010619625@qq.com;dearzppzpp@163.com
作者简介:高铭壑(1995—)，女，吉林长岭人，硕士研究生，主要研究方向为人工智能. E-mail：1010619625@qq.com
基金资助:
中央高校基本科研业务费专项资金(2018MS024);国家自然科学基金资助项目(61305056);吉林省科技发展计划项目(20190303133SF)

Air quality prediction approach based on integrating forecasting dataset

Minghe GAO¹(),Ying ZHANG^1,*(),Rongrong ZHANG¹,Zihao HUANG¹,Linyan HUANG¹,Fanyu LI¹,Xin ZHANG²,Yanhao WANG¹

1. School of Control and Computer Engineering, North China Electric Power University, Beijing 102206, China
2. School of Computer Science and Technology, Changchun University of Science and Technology, Jilin 130022, China

Received:2019-07-18 Online:2020-04-20 Published:2020-04-16
Contact: Ying ZHANG E-mail:1010619625@qq.com;dearzppzpp@163.com
Supported by:
中央高校基本科研业务费专项资金(2018MS024);国家自然科学基金资助项目(61305056);吉林省科技发展计划项目(20190303133SF)

摘要/Abstract

摘要：

采用LightGBM预测模型对空气质量预测问题进行研究,提出并设计一种基于预测性特征的空气质量预测方法,有效地预测北京市区内未来24 h核心表征空气质量的PM_2.5质量浓度。在构建预测方案过程中,分析训练数据集特性开展数据清洗,利用随机森林与线性插值相结合的方法,解决数据大量缺失以及噪声干扰问题;提出使用预测性数据特征方法,同时设计相关统计特征,提高预测结果的准确性;采用滑窗机制挖掘高维时间特征,增加数据特征数量级;对预测模型的工作性能和结果进行详细分析,并结合基线模型进行对比评价。试验结果表明,基于预测性特征结合采用LightGBM预测模型的方案具有更高的预测精度。

关键词: 预测数据融合, 高维统计特征, 空气质量预测, 机器学习

Abstract:

Towarding the air quality prediction research problem, LightGBM was employed to propose and design a predictive feature-based air quality prediction approach, which could effectively predict the PM_2.5 concentration, i.e., the key indicator reflecting air quality, in the upcoming 24-hour within Beijing. During constructing the prediction solution, the features of the training data set was analyzed to execute data cleansing, and the methods of random forest and linear interpolation were used to solve the problem of high data loss and noise interference. The predictive data features were integrated into the dataset, and meanwhile the corresponding statistical features were designed to imiprove the prediction accurancy. The sliding window mechanism was used to mine high-dimensional time features and increase the quantity of data features. The performance and result of the proposed approach were analyzed in details through comparing with the basedline models. The experimental results showed that compared with other model methods, the proposed LightGBM-based prediction approach with integrating forecasting data had higher prediction accuracy.

Key words: predictive data fusion, high dimensional statistical features, air quality prediction, machine learning

中图分类号:

TP18

高铭壑,张莹,张蓉蓉,黄子豪,黄琳焱,李繁菀,张昕,王彦浩. 基于预测数据特征的空气质量预测方法[J]. 山东大学学报 (工学版), 2020, 50(2): 91-99.

Minghe GAO,Ying ZHANG,Rongrong ZHANG,Zihao HUANG,Linyan HUANG,Fanyu LI,Xin ZHANG,Yanhao WANG. Air quality prediction approach based on integrating forecasting dataset[J]. Journal of Shandong University(Engineering Science), 2020, 50(2): 91-99.

图/表 15

图1

图2

表1

图3

图4

图5

表2

图6

图7

图8

表3

表4

图9

图10

表5

参考文献 24

1	HUANG J , DUAN N , JI P , et al. A crowd source-based sensing system for monitoring fine-grained air quality in urban environments[J]. IEEE Internet of Things Journal, 2018, 6 (2): 3240- 3247.
2	LI X , PENG L , HU Y , et al. Deep learning architecture for air quality predictions[J]. Environmental Science and Pollution Research, 2016, 23 (22): 22408- 22417. doi: 10.1007/s11356-016-7812-9
3	ZHOU Q , JIANG H , WANG J , et al. A hybrid model for PM_2.5 forecasting based on ensemble empirical mode decomposition and a general regression neural network[J]. Science of the Total Environment, 2014, 496, 264- 274. doi: 10.1016/j.scitotenv.2014.07.051
4	HOCHREITER S , SCHMIDHUBER J . Long short-term memory[J]. Neural Computation, 1997, 9 (8): 1735- 1780. doi: 10.1162/neco.1997.9.8.1735
5	COSMA A C , SIMHA R . Machine learning method for real-time non-invasive prediction of individual thermal preference in transient conditions[J]. Building and Environment, 2019, 148, 372- 383. doi: 10.1016/j.buildenv.2018.11.017
6	ZHU D , CAI C , YANG T , et al. A machine learning approach for air quality prediction: model regularization and optimization[J]. Big Data and Cognitive Computing, 2018, 2 (1): 1- 15.
7	WANG D , WEI S , LUO H , et al. A novel hybrid model for air quality index forecasting based on two-phase decomposition technique and modified extreme learning machine[J]. Science of the Total Environment, 2017, 580, 719- 733. doi: 10.1016/j.scitotenv.2016.12.018
8	MAHAJAN S , LIU H M , TSAI T C , et al. Improving the accuracy and efficiency of PM_2.5 forecast service using cluster-based hybrid neural network model[J]. IEEE Access, 2018, 6, 19193- 19204. doi: 10.1109/ACCESS.2018.2820164
9	ZHENG Y, YI X, LI M, et al. Forecasting fine-grained air quality based on big data[C]//Proceedings of the 21th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. Sydney, Australia: Associ-ation for Computing Machinery, 2015: 2267-2276.
10	ZHANG C, YUAN D. Fast fine-grained air quality index level prediction using random forest algorithm on cluster computing of spark[C]//Proceeding of 2015 IEEE 12th Intl Conf on Ubiquitous Intelligence and Computing and 2015 IEEE 12th Intl Conf on Autonomic and Trusted Computing and 2015 IEEE 15th Intl Conf on Scalable Computing and Communications and Its Associated Workshops (UIC-ATC-ScalCom). Beijing, China: IEEE, 2015: 929-934.
11	GAO M , YIN L , NING J . Artificial neural network model for ozone concentration estimation and Monte Carlo analysis[J]. Atmospheric Environment, 2018, 184, 129- 139. doi: 10.1016/j.atmosenv.2018.03.027
12	ZHENG Y, LIU F, HSIEH HP. U-air: when urban air quality inference meets big data[C]//Proceedings of the 19th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. Chicago, USA: Association for Computing Machinery, 2013: 1436-1444.
13	HSIEH H P, LIN S D, ZHENG Y. Inferring air quality for station location recommendation based on urban big data[C]//Proceedings of the 21th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. Sydney, Australia: Association for Computing Machinery, 2015: 437-446.
14	WANG J , SONG G . A deep spatial-temporal ensemble model for air quality prediction[J]. Neurocomputing, 2018, 314, 198- 206. doi: 10.1016/j.neucom.2018.06.049
15	HUANG C J , KUO P H . A deep cnn-lstm model for particulate matter (PM_2.5) forecasting in smart cities[J]. Sensors, 2018, 18 (7): 1- 22.
16	SUN W, DUAN N, JI P, et al. Intelligent in-vehicle air quality management: a smart mobility application dealing with air pollution in the traffic[C]//Proceeding of 23rd ITS World Congress. Melbourne, Australia: Intelligent Transport Systems Australia, 2016: 1-12.
17	MA C, DUAN N, SUN W, et al. Reducing air pollution exposure in a road trip[C]//Proceeding of 24rd ITS World Congress. Montreal, Canada: Intelligent Transport Systems Australia, 2017: 1-12.
18	CHENG Y , ZHANG S , HUAN C , et al. Optimization on fresh outdoor air ratio of air conditioning system with stratum ventilation for both targeted indoor air quality and maximal energy saving[J]. Building and Environment, 2019, 147, 11- 22. doi: 10.1016/j.buildenv.2018.10.009
19	SUN W, ZHU J, DUAN N, et al. Moving object map analytics: a framework enabling contextual spatial-temporal analytics of Internet of Things applications[C]//Proceeding of 2016 IEEE International Conference on Service Operations and Logistics, and Informatics (SOLI). Beijing, China: IEEE, 2016: 101-106.
20	ROY S S, PRATYUSH C, BARNA C. Predicting ozone layer concentration using multivariate adaptive regression splines, random forest and classification and regression tree[C]//Proceeding of International Workshop Soft Computing Applications. Arad, Romania: Springer, 2016: 140-152.
21	CHANG J C , HANNA S R . Air quality model performance evaluation[J]. Meteorology and Atmospheric Physics, 2004, 87 (1/2/3): 167- 196.
22	MEIJERING E . A chronology of interpolation: from ancient astronomy to modern signal and image processing[J]. Proceedings of the IEEE, 2002, 90 (3): 319- 342.
23	KE G, MENG Q, FINLEY T, et al. Lightgbm: a highly efficient gradient boosting decision tree[C]//Proceeding of 31st Conference on Neural Information Processing Systems. Long Beach, USA: Curran Associates, Inc., 2017: 3146-3154.
24	FRIEDMAN JH . Greedy function approximation: a gradient boosting machine[J]. Annals of Statistics, 2001, 29 (5): 1189- 1232.

多维度评价

Viewed

Full text

Abstract

Cited

Shared

Discussed

参数	缺失数据量/条	缺失比例/%
PM₁₀	83 263	26.771 8
CO	42 813	13.765 8
O₃	20 421	6.566 0
PM_2.5	20 389	6.555 7
NO₂	18 651	5.996 9
SO₂	18 548	5.963 8

特征	名称	特征描述
时间特征	day_of_week	一周内日期, 1:星期一, 2:星期二, …7:星期日
	day_of_hour	一天内时刻, 00:00—23:00时
	day_of_month	一月内日期序列, 0:1日, 1:2日, …, 30:31日
	isweekend	是否为双休日
	hour_to_predict	要预测的时间, 0~24 h
	CO_1, …, CO_144	历史144 h CO特征
	h_temperature_1, …, h_temperature_144	历史144 h气温特征
气象特征	humity	给定时间内的湿度
	weather	给定时间内的天气状况
	temperature	给定时间内的温度
	wind_direction	给定时间内的风向
	wind_speed	给定时间内的风速
	pressure	给定时间内的气压
空气质量特征	CO	给定时间内的CO质量浓度
	PM₁₀	给定时间内的PM₁₀质量浓度
	NO₂	给定时间内的NO₂质量浓度
	O₃	给定时间内的O₃质量浓度
	SO₂	给定时间内的SO₂质量浓度
天气预报特征	temperature_1, …, temperature_24	未来24 h温度
	humity_1, …, humity_24	未来24 h湿度
	pressure_1, …, pressure_24	未来24 h气压
	weather_1, …, weather_24	未来24 h天气状况
	wind_direction_1, …, wind_direction24	未来24 h风向
	wind_speed_1, …, wind_speed_24	未来24 h风速
统计特征	mean_pm25\PM10\O3_1, mean_pm25\PM10\O3_3, mean_pm25\PM10\O3_5	前1、3、5 d PM_2.5\PM₁₀\O₃平均质量浓度
	max_pm25\PM10\O3_1, max_pm25\PM10\O3_3, max_pm25\PM10\O3_5	前1、3、5 d PM_2.5\PM₁₀\O₃最大质量浓度
	min_pm25\PM10\O3_1, min_pm25\PM10\O3_3, min_pm25\PM10\O3_5	前1、3、5 d的PM_2.5\PM₁₀\O₃最小质量浓度
	pm25_13\O3_13\pm10_13	PM_2.5\O₃\PM₁₀前1 d与前3 d平均质量浓度比值
	pm25_35\O3_35\pm10_35	PM_2.5\O₃\PM₁₀前3 d与前5 d平均质量浓度比值

项目	PM_2.5	PM₁₀	NO₂	CO	O₃	SO₂
平均	58.8	88.1	45.8	1.0	55.7	8.98
标准	66.1	89.3	32.1	1.0	53.8	11.7
最小	2.0	5.0	1.0	0.1	1.0	1.0
25%	16.0	37.0	20.0	0.4	29.1	2.0
50%	39.0	70.0	39.0	0.7	45.0	5.0
75%	77.0	113.0	66.0	1.2	79.0	11.0
最大	1 004.0	3 000.0	300.0	15.0	504.0	307.0
条目数	290 621	227 747	292 359	268 197	290 589	292 462

项目	温度/℃	气压/ hPa	湿度/ %	风向/ (°)	风速/ (m·s^-1)
平均	38.2	1 026.8	37.1	35 487.5	9.8
标准差	5 030.6	5 025.7	18.9	184 454.8	5.5
最小	-21.3	940.0	5.0	0.0	0.1
25%	2.5	994.2	23.0	78.0	5.6
50%	13.8	1 005.6	33.0	48.0	8.5
75%	23.2	1 016.9	48.0	280.0	12.9
最大	999 999.0	999 999.0	100.0	999 999.0	30.0
条目数	15 8047	15 8047	15 8047	157 813	157 813

方法模型	S	M	A
XGBoost	0.430 7	33.094 8	27.054 5
GBDT	0.432 9	33.306 0	27.263 8
本文方法	0.422 9	32.871 1	26.436 0
DNN	0.540 6	42.515 2	33.436 5
LightGBM(无预报)	0.429 8	33.892 3	26.682 5

基于预测数据特征的空气质量预测方法

Air quality prediction approach based on integrating forecasting dataset

RichHTML

PDF (PC)

摘要/Abstract

引用本文

使用本文

图/表 15

参考文献 24

相关文章 15

多维度评价

本文评价

推荐阅读 10

[1]	祝明,石承龙,吕潘,刘现荣,孙驰,陈建城,范宏运. 基于优化长短时记忆网络的深基坑变形预测方法及其工程应用[J]. 山东大学学报 (工学版), 2025, 55(3): 141-148.
[2]	常新功,苏敏惠,周志刚. 基于进化集成的图神经网络解释方法[J]. 山东大学学报 (工学版), 2024, 54(4): 1-12.
[3]	乔慧妍,段学龙,解驰皓,赵冬慧,马玉玲. 基于异常点检测的心理健康辅助诊断方法[J]. 山东大学学报 (工学版), 2024, 54(4): 76-85.
[4]	刘新,刘冬兰,付婷,王勇,常英贤,姚洪磊,罗昕,王睿,张昊. 基于联邦学习的时间序列预测算法[J]. 山东大学学报 (工学版), 2024, 54(3): 55-63.
[5]	岳仁峰,张嘉琦,刘勇,范学忠,李琮琮,孔令鑫. 基于颜色和纹理特征的立体车库锈蚀检测技术[J]. 山东大学学报 (工学版), 2024, 54(3): 64-69.
[6]	陈成,董永权,贾瑞,刘源. 基于交互序列特征相关性的可解释知识追踪[J]. 山东大学学报 (工学版), 2024, 54(1): 100-108.
[7]	卞小曼,王小琴,蓝如师,刘振丙,罗笑南. 基于相似性保持和判别性分析的快速视频哈希算法[J]. 山东大学学报 (工学版), 2023, 53(6): 63-69.
[8]	李鸿钊,张庆松,刘人太,陈新,辛勤,石乐乐. 浅埋地铁车站施工期地表变形风险预警[J]. 山东大学学报 (工学版), 2023, 53(6): 82-91.
[9]	袁高腾,周晓峰,郭宏乐. 基于特征选择算法的ECG信号分类[J]. 山东大学学报 (工学版), 2022, 52(4): 38-44.
[10]	聂秀山,马玉玲,乔慧妍,郭杰,崔超然,于志云,刘兴波,尹义龙. 任务粒度视角下的学生成绩预测研究综述[J]. 山东大学学报 (工学版), 2022, 52(2): 1-14.
[11]	孙鸿昌,周风余,单明珠,翟文文,牛兰强. 基于模式划分的空调能耗混合填补方法[J]. 山东大学学报 (工学版), 2022, 52(1): 9-18.
[12]	袁高腾,刘毅慧,黄伟,胡兵. 基于Gabor特征的乳腺肿瘤MR图像分类识别模型[J]. 山东大学学报 (工学版), 2020, 50(3): 15-23.
[13]	张大鹏,刘雅军,张伟,沈芬,杨建盛. 基于异质集成学习的虚假评论检测[J]. 山东大学学报 (工学版), 2020, 50(2): 1-9.
[14]	刘玉田, 孙润稼, 王洪涛, 顾雪平. 人工智能在电力系统恢复中的应用综述[J]. 山东大学学报 (工学版), 2019, 49(5): 1-8.
[15]	李童,马然,郑鸿鹤,安平,胡翔宇. 基于视频统计特征的差错敏感度模型[J]. 山东大学学报 (工学版), 2019, 49(2): 116-121.