Please wait a minute...
 
自然资源遥感  2025, Vol. 37 Issue (4): 173-183    DOI: 10.6046/zrzyyg.2024146
  技术应用 本期目录 | 过刊浏览 | 高级检索 |
高斯混合模型及其在工业热源遥感识别中的应用
李乐林1,2(), 王文曦3, 杨文涛1, 陈浩2, 彭焕华2, 赵茜3
1.湖南科技大学地理空间信息技术国家地方联合工程实验室,湘潭 411201
2.湖南科技大学区域可持续发展研究院,湘潭 411201
3.湖南科技大学地球科学与空间信息工程学院,湘潭 411201
Gaussian mixture model and its application in remote sensing identification of industrial heat sources
LI Lelin1,2(), WANG Wenxi3, YANG Wentao1, CHEN Hao2, PENG Huanhua2, ZHAO Qian3
1. National-Local Joint Engineering Laboratory of Geo-Spatial Information Technology, Hunan University of Science and Technology, Xiangtan 411201, China
2. Institute for Local Sustainable Development Goals, Hunan University of Science and Technology, Xiangtan 411201, China
3. School of Earth Sciences and Spatial Information Engineering, Hunan University of Science and Technology, Xiangtan 411201, China
全文: PDF(5032 KB)   HTML  
输出: BibTeX | EndNote (RIS)      
摘要 

工业热源活动的精细提取是我国大气污染防治和工业经济预测的重要前提,然而因热源特征不明、类型判定不准等问题,导致工业热源遥感监测目前难以大范围推广应用。该文选取湖南省作为研究区,基于2015—2021年Suomi-NPP VIIRS夜火数据,结合空间滤波与具有噪声的基于分层密度空间聚类(hierarchical density-based spatial clustering of applications with noise,HDBSCAN)算法提取夜间工业热源,利用高斯混合模型构建不同类别的工业热源温度特征模板函数,根据相同类别的温度相似性判断工业热源子类别,总体分类精度达86.31%。并分析湖南省工业热源布局情况,结果表明: ①全省工业以石化厂为主,煤化厂数量最少; ②冶金企业的热辐射强度最大,集中在娄底市-湘潭市-株洲市一带; ③2015—2019年间,全省工业热源数量呈减少趋势,说明政府部门在“十三五”期间对省内“散乱污”工厂进行有效整治,新冠疫情期间热源数量整体变化不大,在政府有效调控下已逐步复工复产; ④通过分析热辐射排放强度与能源消耗量、工业污染物排放的相关指标的灰色关联度,以工业综合能源消费量、工业二氧化硫排放量和热辐射排放强度为标准,探究湖南省“能耗-污染-热排”的相关情况,并将各市州划分为7种类型。研究可为地方政府动态监测当地重点工业企业的生产活动状况提供信息来源与数据支持,摸清该地区不同产业类型的工业企业空间分布格局及演变趋势,有利于政府以及相关部门制定产业转型政策,践行可持续发展理念。

服务
把本文推荐给朋友
加入引用管理器
E-mail Alert
RSS
作者相关文章
李乐林
王文曦
杨文涛
陈浩
彭焕华
赵茜
关键词 Suomi-NPP VIIRS夜火数据工业热源温度特征高斯混合模型遥感识别    
Abstract

Precisely extracting the information of industrial heat source activities serves as a significant prerequisite for the prevention and control of air pollution and the prediction of industrial economy in China. However, due to unclear heat source characteristics and inaccurate type determination, the remote sensing monitoring of industrial heat sources fails to be widely applied. This study investigated Hunan Province based on the Suomi-NPP VIIRS Nightfire data from 2015 to 2021. First, this study extracted nighttime industrial heat sources from the data using spatial filtering and the hierarchical density-based spatial clustering of applications with noise (HDBSCAN) algorithm. Second, this study constructed temperature characteristic template functions for different industrial heat sources using the Gaussian mixture model. Third, this study determined the subcategories of industrial heat sources according to temperature similarity of the same categories, achieving an overall classification accuracy of 86.31 %. Finally, this study obtained the layout of industrial heat sources in Hunan Province. The results indicate that the industry in Hunan Province was dominated by petrochemical plants, with the smallest number of coal-to-chemical plants. Metallurgical enterprises, showing the highest heat radiation intensity, were primarily distributed in the Loudi-Xiangtan-Zhuzhou area. From 2015 to 2019, industrial heat sources in Hunan Province showed a decreasing trend, indicating that relevant government departments effectively rectified the scattered, non-compliant, and polluting factories in Hunan Province during the 13th Five-Year Plan period. During the COVID-19 pandemic, the number of heat sources changed slightly since work and production were gradually resumed under the effective regulation of the government. This study analyzed the grey relational degrees between the heat radiation emission intensity and the relevant indicators of energy consumption and industrial pollutant emissions. Based on the comprehensive industrial energy consumption, industrial sulfur dioxide emissions, and heat radiation emission intensity, this study explored the relevant situation of energy consumption, pollution, and heat emissions in Hunan Province, dividing the cities and prefectures into seven types accordingly. Overall, this study provides information sources and data support for local governments to dynamically monitor the production activities of local key industrial enterprises. Ascertaining the spatial distribution patterns and evolutionary trends of different industrial enterprises will contribute to the formulation of industrial transformation policies by the government and relevant departments and the practice of sustainable development.

Key wordsSuomi-NPP VIIRS Nightfire data    industrial heat source    temperature characteristics    Gaussian mixture model    remote sensing identification
收稿日期: 2024-04-19      出版日期: 2025-09-03
ZTFLH:  TP79  
基金资助:国家科技基础资源调查专项项目“全球地表覆盖时空变化数字地图编研”(2019FY202502);湖南省自然科学基金项目“后扶贫时代基于多源遥感与深度学习的乡村防返贫动态监测关键技术研究”(2023JJ30232);与国家级大学生创新训练计划项目“面向乡村振兴的融合多源遥感数据返贫监测研究”(202210534038)
作者简介: 李乐林(1981-),男,博士,副教授,主要从事多源卫星遥感及其应用研究。Email: lilelin@hnust.edu.cn
引用本文:   
李乐林, 王文曦, 杨文涛, 陈浩, 彭焕华, 赵茜. 高斯混合模型及其在工业热源遥感识别中的应用[J]. 自然资源遥感, 2025, 37(4): 173-183.
LI Lelin, WANG Wenxi, YANG Wentao, CHEN Hao, PENG Huanhua, ZHAO Qian. Gaussian mixture model and its application in remote sensing identification of industrial heat sources. Remote Sensing for Natural Resources, 2025, 37(4): 173-183.
链接本文:  
https://www.gtzyyg.com/CN/10.6046/zrzyyg.2024146      或      https://www.gtzyyg.com/CN/Y2025/V37/I4/173
Fig.1  基于VNF数据提取的工业热源分布情况
Fig.2  不同类别工业热源温度特征
企业类别 GMM模型 R2 调整后R2 RMSE/%
冶金 f ( x ) = 4.55 e - x - 874.2 85.41 2 + 3.315 e - x - 1094 1443 2 + 0.2803 e - x - 1531 1302 2 0.993 7 0.993 2 0.114 1
水泥 f ( x ) = 11.52 e - x - 840 93.58 2 0.937 2 0.934 8 0.870 4
石化 f ( x ) = 4.894 e - x - 1191 168 2 + 2.275 e - x - 858.4 119.6 2 + 0.1839 e - x - 1586 206.1 2 0.988 0 0.986 9 0.218 5
煤化 f ( x ) = 4.533 e - x - 1752 103.8 2 + 2.041 e - x - 1776 324.1 2 0.970 9 0.967 6 0.277 8
Tab.1  工业热源子类温度特征GMM模型
分类类别 参考类别 用户精度/%
冶金 水泥 石化 煤化
冶金 34 2 3 1 85.00
水泥 2 38 3 2 84.44
石化 3 4 64 0 90.14
煤化 0 1 2 9 75.00
制图精度/% 89.47 84.44 88.89 75.00
总体精度/% 86.31
Tab.2  工业热源分类精度
Fig.3  研究区各类别工业热源分布
市级行政区 冶金 水泥 石化 煤化 共计
长沙市 4 8 6 0 18
株洲市 2 2 6 0 10
湘潭市 5 5 7 1 18
衡阳市 9 1 9 2 21
邵阳市 0 3 4 1 8
岳阳市 0 5 11 1 17
常德市 1 6 4 0 11
张家界市 0 2 0 0 2
益阳市 0 2 2 1 5
郴州市 8 1 11 1 21
永州市 1 2 4 0 7
怀化市 1 2 4 0 7
娄底市 6 5 3 4 18
湘西土家族苗族自治州 2 1 1 1 5
共计 39 45 72 12 168
Tab.3  各市工业热源数量统计
Fig.4  重点经济区工业热源标准距离圆分析
Fig.5-1  研究区各类别工业热源热排放空间密度分布
Fig.5-2  研究区各类别工业热源热排放空间密度分布
Fig.6  研究区各年份工业热源变化趋势
评价项 关联度 排名
X8 0.846 1
X3 0.788 2
X5 0.761 3
X7 0.756 4
X1 0.726 5
X4 0.712 6
X6 0.704 7
X2 0.700 8
Tab.4  灰色关联度结果
[1] 李宇军. 生态文明视野下的企业环保责任——“腾格里沙漠违法排污” 案例分析[J]. 杭州师范大学学报(社会科学版), 2015, 37(6):119-125.
Li Y J. The enterprise environmental responsibility from the perspective of ecological civilization:A case study on illegal pollution discharge in Tengger Desert[J]. Journal of Hangzhou Normal University (Humanities and Social Sciences), 2015, 37(6):119-125.
[2] 何慧妍, 杨庆媛, 毕国华, 等. 重庆市2009—2018年工业用地供应规模时空特征分析[J]. 长江流域资源与环境, 2021, 30(4):808-817.
He H Y, Yang Q Y, Bi G H, et al. Spatio-temporal characteristics of industrial land supply in Chongqing from 2009 to 2018[J]. Resources and Environment in the Yangtze Basin, 2021, 30(4):808-817.
[3] 高菠阳, 罗会琳, 黄志基, 等. 中国工业用地出让价格空间格局及影响因素[J]. 地球信息科学学报, 2020, 22(6):1189-1201.
Gao B Y, Luo H L, Huang Z J, et al. Research on the spatial layout of and factors affecting the price of industrial land in China[J]. Journal of Geo-Information Science, 2020, 22(6):1189-1201.
[4] 史焱文, 李小建, 孟德友. 典型农区县域工业化进程评价及时空演进——以河南省为例[J]. 经济地理, 2020, 40(10):118-126,153.
Shi Y W, Li X J, Meng D Y. Evaluation of county industrialization process and analysis of spatial and temporal evolution in typical agricultural areas:A case study of Henan Province[J]. Economic Geography, 2020, 40(10):118-126,153.
[5] 傅国伟. 当代环境规划的定义、作用与特征分析[J]. 中国环境科学, 1999, 19(1):72-76.
Fu G W. Definition,function and characteristics analysis of modern environmental planning[J]. China Environmental Science, 1999, 19(1):72-76.
[6] 任胜钢, 张如波, 袁宝龙. 长江经济带工业生态效率评价及区域差异研究[J]. 生态学报, 2018, 38(15):5485-5497.
Ren S G, Zhang R B, Yuan B L. Industrial eco-efficiency evaluation and regional differences of Yangtze River Economic Belt[J]. Acta Ecologica Sinica, 2018, 38(15):5485-5497.
[7] 赖建波. 重工业热源的遥感识别及空间分布格局研究[D]. 兰州: 西北师范大学, 2020.
Lai J B. Study on remote sensing identification and spatial distribution pattern of heat source in heavy industry[D]. Lanzhou: Northwest Normal University, 2020.
[8] 陈鹏飞, 卢力, 朱华忠, 等. 基于不同分辨率影像的水泥厂遥感识别适宜性研究[J]. 环境污染与防治, 2015, 37(9):20-28.
Chen P F, Lu L, Zhu H Z, et al. Research on the suitability of different resolutions of image for identification of the cement enterprise using remote sensing[J]. Environmental Pollution and Control, 2015, 37(9):20-28.
[9] 陈鹏飞, 卢力, 朱华忠, 等. 不同分辨率遥感影像的钢铁厂识别适宜性研究[J]. 地球信息科学学报, 2015, 17(9):1119-1127.
Chen P F, Lu L, Zhu H Z, et al. Research on the suitability of image at different resolutions for the identification of steel enterprise using remote sensing[J]. Journal of Geo-Information Science, 2015, 17(9):1119-1127.
[10] 于一凡, 潘军, 邢立新, 等. 短波红外波段高温目标识别的可行性分析[J]. 国土资源遥感, 2014, 26(1):25-30.doi:10.6046/gtzyyg.2014.01.05.
Yu Y F, Pan J, Xing L X, et al. Feasibility analysis of shortwave infrared band for recognition of high temperature target[J]. Remote Sensing for Land and Resources, 2014, 26(1):25-30.doi:10.6046/gtzyyg.2014.01.05.
[11] Zhou Y, Zhao F, Wang S X, et al. A method for monitoring iron and steel factory economic activity based on satellites[J]. Sustainability, 2018, 10(6):1935.
[12] 孙爽, 姜磊, 刘保献, 等. 中高分辨率卫星影像的装置级别工业热源识别[J]. 遥感学报, 2024, 28(8):2101-2112.
Sun S, Jiang L, Liu B X, et al. Method development for device-le-vel industrial heat source identification using medium and high-reso-lution satellite images[J]. National Remote Sensing Bulletin, 2024, 28(8):2101-2112.
[13] Elvidge C D, Zhizhin M, Baugh K, et al. Methods for global survey of natural gas flaring from visible infrared imaging radiometer suite data[J]. Energies, 2016, 9(1):14.
[14] 孙佳琪, 刘永学, 董雁伫, 等. 基于Suomi-NPP VIIRS夜间热异常产品的城市工业热源分类——以京津冀地区为例[J]. 地理与地理信息科学, 2018, 34(3):13-19.
Sun J Q, Liu Y X, Dong Y Z, et al. Classification of urban industrial heat sources based on Suomi-NPP VIIRS nighttime thermal anomaly products:A case study of the Beijing-Tianjin-Hebei Region[J]. Geography and Geo-Information Science, 2018, 34(3):13-19.
[15] 李博, 范俊甫, 韩留生, 等. 基于VIIRS Nightfire数据的山东省工业热源分类提取与变化特征分析[J]. 遥感技术与应用, 2022, 37(4):919-928.
Li B, Fan J F, Han L S, et al. Classification extraction and variation characteristics analysis of industrial heat sources in Shandong Province based on VIIRS nightfire data[J]. Remote Sensing Technology and Application, 2022, 37(4):919-928.
[16] 张钦挺, 邹滨, 刘宁, 等. 耦合温度特征的工业热源ANN遥感识别与时空演化分析[J]. 遥感学报, 2024, 28(4):956-968.
Zhang Q T, Zou B, Liu N, et al. Satellite-based ANN identification and spatiotemporal evolution analysis of industrial heat sources coupled with temperature characteristics[J]. National Remote Sensing Bulletin, 2024, 28(4):956-968.
[17] Ma Y, Ma C H, Liu P, et al. Spatial-temporal distribution analysis of industrial heat sources in the US with geocoded,tree-based,large-scale clustering[J]. Remote Sensing, 2020, 12(18):3069.
[18] Ma C H, Niu Z, Ma Y, et al. Assessing the distribution of heavy industrial heat sources in India between 2012 and 2018[J]. ISPRS International Journal of Geo-Information, 2019, 8(12):568.
[19] 纪轩禹. 京津冀地区工业热源活动卫星遥感监测及环境影响研究[D]. 青岛: 山东科技大学, 2019.
Ji X Y. Evaluation of the industrial heat sources activities over Jing-Jin-Ji region and their impact on the air pollution based on satellite data[D]. Qingdao: Shandong University of Science and Technology, 2019.
[20] 陶金花, 范萌, 顾坚斌, 等. 新冠病毒疫情期间复工复产卫星遥感监测[J]. 遥感学报, 2020, 24(7):824-836.
Tao J H, Fan M, Gu J B, et al. Satellite observations of the return-to-work over China during the period of COVID-19[J]. Journal of Remote Sensing, 2020, 24(7):824-836.
[21] 孙爽, 李令军, 赵文吉, 等. 基于热异常遥感的冀南城市群工业能耗及大气污染[J]. 中国环境科学, 2019, 39(7):3120-3129.
Sun S, Li L J, Zhao W J, et al. Industrial pollution emissions based on thermal anomaly remote sensing monitoring:A case study of southern Hebei urban agglomerations,China[J]. China Environmental Science, 2019, 39(7):3120-3129.
[22] 郝丽春, 孟庆岩, 葛小三, 等. 一种基于八分位法的工业热污染区提取方法[J]. 遥感技术与应用, 2020, 35(2):469-477.
Hao L C, Meng Q Y, Ge X S, et al. Extraction method of industrial heat pollution area based on octave method[J]. Remote Sensing Technology and Application, 2020, 35(2):469-477.
[23] Elvidge C D, Zhizhin M, Hsu F C, et al. VIIRS nightfire:Satellite pyrometry at night[J]. Remote Sensing, 2013, 5(9):4423-4449.
[24] Elvidge C D, Zhizhin M, Baugh K, et al. Extending nighttime combustion source detection limits with short wavelength VIIRS data[J]. Remote Sensing, 2019, 11(4):395.
[25] Cherifa S, Messaoud R. New technique to use the GMM in speaker recognition system (SRS)[C]// 2013 International Conference on Computer Applications Technology (ICCAT).IEEE, 2013: 1-5.
[26] Silverman B W. Density estimation in action[M]// Density Estimation for Statistics and Data Analysis.London:Routledge, 2018: 120-158.
[27] 蒋姝睿, 王玥, 王萌, 等. 区域视角下中国工业行业与工业污染关系[J]. 中国环境科学, 2017, 37(11):4380-4387.
Jiang S R, Wang Y, Wang M, et al. Industrial sectors and pollution in China based on the regional perspective[J]. China Environmental Science, 2017, 37(11):4380-4387.
[1] 邹海靖, 邹滨, 王玉龙, 张波, 邹伦文. 基于多尺度样本集优化策略的矿区工业固废及露天采场遥感识别[J]. 自然资源遥感, 2025, 37(3): 1-8.
[2] 阳驰轶, 官海翔, 吴玮, 刘美玉, 李颖, 苏伟. 基于国产GF-3雷达影像的农田洪涝遥感监测方法[J]. 自然资源遥感, 2023, 35(4): 71-80.
[3] 王丽英, 马旭伟, 有泽, 王世超, CAMARA Mahamadou. 基于多元GMM的机载多光谱LiDAR点云空谱联合分类[J]. 自然资源遥感, 2023, 35(3): 88-96.
[4] 张成业, 邢江河, 李军, 桑潇. 基于U-Net网络和GF-6影像的尾矿库空间范围识别[J]. 自然资源遥感, 2021, 33(4): 252-257.
[5] 闫驰, 焦润成, 曹颖, 南赟, 王晟宇, 郭学飞. 无人机倾斜摄影在泥石流灾害识别分析中的应用——以北京房山区史家营曹家坊泥石流为例[J]. 国土资源遥感, 2020, 32(4): 251-257.
[6] 胡波, 朱谷昌, 张远飞, 冷超. 基于高斯混合模型的遥感信息提取方法研究[J]. 国土资源遥感, 2012, 24(4): 41-47.
[7] 林婷, 刘湘南, 谭正. 基于ICA和高光谱指数的水稻Zn污染监测模型[J]. 国土资源遥感, 2011, 23(2): 59-64.
[8] 钟仕全, 陈燕丽, 莫建飞, 陈玉兰. 基于中巴02B星遥感数据的油菜识别技术研究[J]. 国土资源遥感, 2010, 22(2): 77-79.
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
京ICP备05055290号-2
版权所有 © 2015 《自然资源遥感》编辑部
地址:北京学院路31号中国国土资源航空物探遥感中心 邮编:100083
电话:010-62060291/62060292 E-mail:zrzyyg@163.com
本系统由北京玛格泰克科技发展有限公司设计开发