| 戴剑勇,汪恒浩.基于可视图的衡阳市大气污染物时间序列分析[J].南华大学学报(自然科学版),2021,(3):30~35.[DAI Jianyong,WANG Henghao.Visibility Graphs Analysis of Air Pollutants Time Series in Hengyang[J].Journal of University of South China(Science and Technology),2021,(3):30~35.] |
| 基于可视图的衡阳市大气污染物时间序列分析 |
| Visibility Graphs Analysis of Air Pollutants Time Series in Hengyang |
| 投稿时间：2021-01-04 |
| DOI： |
| 中文关键词: 大气污染物 时间序列 复杂网络可视图 拓扑特性 |
| 英文关键词:atmospheric pollutants time series complex networks viewable groph topological characteristics |
| 基金项目:湖南省教育厅重点资助科研项目(18A235);铀矿冶放射性控制技术湖南省工程研究中心&湖南省铀尾矿库退役治理技术工程技术研究中心联合开放重点课题(2018YKZX1001) |
|
| 摘要点击次数: 539 |
| 全文下载次数: 465 |
| 中文摘要: |
| 为研究大气污染物时间序列的非线性特征,基于2016年1月至2017年12月衡阳市PM2.5和PM10质量浓度时间序列,应用可视图方法将两组时间序列映射到复杂网络中,并研究了相应网络的拓扑性质。结果表明:PM2.5和PM10浓度时间序列网络的平均聚类系数、直径、网络密度和平均路径长度等网络特征参数和统计特征参数基本相同。PM2.5浓度时间序列在局部变化趋于平缓,矩阵图中存在明显聚集现象;PM10浓度时间序列存在多个局部峰值,矩阵图中表现为多个重叠的大正方形。两个时间序列网络的累积度分布都具有幂律特征,网络具有无标度特性。原时间序列具有分形序列特征,赫斯特指数均大于0.5,都是持续性时间序列,表明大气污染物浓度时间序列具有长程记忆性。 |
| 英文摘要: |
| In order to study the nonlinear characteristics of atmospheric pollutant time series, based on the PM2.5 and PM10 mass concentration time series in Hengyang City from January 2016 to December 2017, the two sets of time series were mapped to complex networks by adopting the visibility graph algorithm, and the topological properties of the corresponding networks were studied. The results showed that statistical feature parameters and network characteristics, the average clustering coefficient, network diameter, network density and average shortest path length of PM2.5 and PM10 concentration time series networks remained at the same level. The PM2.5 concentration time series tended to flatten the local variation, there was obvious aggregation in the matrix graph, and the PM10 concentration time series had several local peaks, which were represented by several overlapping large squares in the matrix graph. The cumulative degree distribution of both time series networks had power-law tails and thus the network had the characteristics of scale-free. The original time series had the characteristics of the singular sequence, and their Hurst exponent was greater than 0.5, which were continuous time series, indicating that the atmospheric pollutant concentration time series had long-range memory. |
| 查看全文 查看/发表评论 下载PDF阅读器 |
| 关闭 |
|
|
|