北京城市代谢预测研究——基于长短期记忆神经网络模型

doi:10.18306/dlkxjz.2019.06.006

地理科学进展 ›› 2019, Vol. 38 ›› Issue (6): 851-860.doi: 10.18306/dlkxjz.2019.06.006

北京城市代谢预测研究——基于长短期记忆神经网络模型

刘炳春¹(), 齐鑫¹, 王庆山^2,^*()

1. 天津理工大学管理学院,天津 300384
2. 天津农学院,天津 300384

收稿日期:2018-10-04 修回日期:2019-03-18 出版日期:2019-06-28 发布日期:2019-06-27
通讯作者: 王庆山
作者简介:
第一作者简介：刘炳春(1980— ),男,河北衡水人,副教授,主要研究方向为深度学习与数据挖掘、循环经济。E-mail: tjutlbc@163.com
基金资助:
国家自然科学基金项目(71503180);天津市教委社会科学重大项目(2017JWZD16)

Urban metabolism prediction of Beijing City based on long short-term memory neural network

Bingchun LIU¹(), Xin QI¹, Qingshan WANG^2,^*()

1. School of Management, Tianjin University of Technology, Tianjin 300384, China
2. Tianjin Agricultural University, Tianjin 300384, China

Received:2018-10-04 Revised:2019-03-18 Online:2019-06-28 Published:2019-06-27
Contact: Qingshan WANG
Supported by:
National Natural Science Foundation of China, No. 71503180;Major Social Science Projects of Tianjin Education Commission, No. 2017JWZD16.

摘要/Abstract

摘要：

城市化进程提升促使城市环境污染加剧、能源消耗激增、人口密度过大等问题的深层次原因在于城市代谢失调。为精准预测北京市城市代谢变化趋势,论文通过能源消费量及人类活动时间指标测算了1980—2016年北京市体外能代谢率,表征城市代谢程度。据此运用长短期记忆神经网络模型(LSTM)预测了2017—2022年北京各部门体外能代谢率。结果表明：① 基于长短期记忆神经网络的城市代谢预测模型精度较高,能够对北京各部门体外能代谢率进行更为精准的预测;② 2017—2022年间,北京第一产业和总体外能代谢率呈下降趋势,其中第一产业在2017年达到峰值,第二、第三产业及生活部门体外能代谢率将呈现增长趋势。③ 除第一、第三产业和总体外能代谢率外,历史变化的时间扰动幅度先小后大。④ 对各部门体外能代谢率EMR_T的影响贡献度最大的因子为第二产业体外能代谢率EMR₂,最小的为第一产业体外能代谢率EMR₁。论文研究结果可为政策制定者优化城市管理方案、提升城市综合实力提供理论依据和决策支持。

关键词: 长短期记忆神经网络, 体外能代谢率, 城市代谢, 北京

Abstract:

The underlying causes of aggravating urban environmental pollution, escalating energy consumption, population overcrowding, and other urban environmental problems are imbalances in urban metabolism. In order to accurately predict the trend of urban metabolism changes in Beijing City, the exosomatic metabolic rate of Beijing from 1980 to 2016 was estimated by the indicators of energy consumption and human activity time, and the degree of urban metabolism was characterized. Based on the results, the long short-term memory (LSTM) neural network model was used to predict the exosomatic metabolic rate of various sectors in Beijing from 2017 to 2022. The results show that: 1) the urban metabolic prediction model based on LSTM neural network has high accuracy and can make more accurate prediction on the exosomatic metabolic rate of various sectors in Beijing. 2) From 2017 to 2022, the exosomatic metabolic rate of the primary industry and the overall external energy in Beijing show a downward trend, among which the primary industry reached its peak in 2017, and the exosomatic metabolism rates of the secondary and tertiary industries show an increasing trend. 3) Except for the primary industry, tertiary industry and the overall exosomatic metabolic rate, the temporal perturbation of historical change ranged from small to large. 4) The factors that contribute the most to EMR_T are EMR₂, and the least are EMR₁. This study may provide a theoretical basis and decision-making support for policymakers to optimize urban management plans and enhance urban comprehensive strength.

Key words: long short-term memory neural network, exosomatic metabolic rate, urban metabolism, Beijing

刘炳春, 齐鑫, 王庆山. 北京城市代谢预测研究——基于长短期记忆神经网络模型[J]. 地理科学进展, 2019, 38(6): 851-860.

Bingchun LIU, Xin QI, Qingshan WANG. Urban metabolism prediction of Beijing City based on long short-term memory neural network[J]. PROGRESS IN GEOGRAPHY, 2019, 38(6): 851-860.

图/表 10

图1

图2

图3

图4

表1

表2

表3

表4

表5

图5

参考文献 40

[1]	李恒. 2019. 人口集中、城市群对经济增长作用的实证分析: 以中国十大城市群为例[J]. 河南大学学报(社会科学版), 59(1): 43-52.
	[Li H.2019. Empirical analysis of the effects of population concentration and urban agglomeration on economic growth. Journal of Henan University (Social Science), 59(1): 43-52. ]
[2]	李旋旗, 花利忠. 2012. 基于系统动力学的城市住区形态变迁对城市代谢效率的影响[J]. 生态学报, 32(10): 2965-2974. doi: 10.5846/stxb201104250542
	[Li X Q, Hua L Z.2012. Landscape aesthetic assessment based on experiential paradigm assessment technology. Acta Ecologica Sinica, 32(10): 2965-2974. ] doi: 10.5846/stxb201104250542
[3]	刘晔, 耿涌, 赵恒心. 2011. 基于MSIASM方法的中国省级行政区体外能代谢分析[J]. 生态学报, 31(11): 3133-3142.
	[Liu Y, Geng Y, Zhao H X, 2011. Societal metabolism for chinese provinces based on multi-scale integrated analysis of societal metabolism (MSIASM). Acta Ecologica Sinica, 31(11): 43-52. ]
[4]	宋涛, 蔡建明, 倪攀, 等. 2013a. 城市新陈代谢研究综述及展望[J]. 地理科学进展, 32(11): 1650-1661.
	[Song T, Cai J M, Ni P, et al.2013. Review and prospects of urban metabolism research. Progress in Geography, 32(11): 1650-1661. ]
[5]	宋涛, 蔡建明, 倪攀, 等. 2013b. 基于能值和DEA的中国城市新陈代谢效率分析[J]. 资源科学, 35(11): 2166-2173.
	[Song T, Cai J M, Pan N, et al.2013. Chinese urban metabolic efficiencies based on emergy and DEA. Resources Science, 35(11): 2166-2173. ]
[6]	王天送. 2008. 社会代谢多尺度综合评估(MSIASM)的基本理论与实践[J]. 地球科学进展, 27(1): 63-70.
	[Wang T.2008. The rationale of multiple-scale integrated assessment of societal metabolism. Advances in Earth Science, 27(1): 63-70. ]
[7]	王劲峰,徐成东. 2017. 地理探测器: 原理与展望[J]. 地理学报, 72(1): 116-134.
	[Wang J, Xu C.2017. Geodetector: Principle and prospective. Acta Geographica Sinica, 72(1): 116-134. ]
[8]	文琦,施琳娜,马彩虹, 等. 2018. 黄土高原村域多维贫困空间异质性研究: 以宁夏彭阳县为例[J]. 地理学报, 73(10): 1850-1864.
	[Wen Q, Shi L, Ma C, et al.2018. Spatial heterogeneity of multidimensional poverty at the village level: Loess Plateau. Acta Geographica Sinica, 73(10): 1850-1864. ]
[9]	湛东升,张文忠,余建辉, 等. 2015. 基于地理探测器的北京市居民宜居满意度影响机理[J]. 地理科学进展, 34(8): 966-975. doi: 10.18306/dlkxjz.2015.08.004
	[Zhan D, Zhang W, Yu J, et al.2015. Analysis of influencing mechanism of residents' livability satisfaction in Beijing using geographical detector. Progress in Geography, 34(8): 966-975. ] doi: 10.18306/dlkxjz.2015.08.004
[10]	Abdel-Nasser M, Mahmoud K.2017. Accurate photovoltaic power forecasting models using deep LSTM-RNN[J]. Neural Computing & Applications, 10: 1-14.
[11]	Elliot T, Rugani B, Almenar J B, et al.2018. A proposal to integrate system dynamics and carbon metabolism for urban planning[J]. Procedia CIRP, 69: 78-82. doi: 10.1016/j.procir.2017.10.003
[12]	Geng Y, Liu Y, Liu D, et al.2011. Regional societal and ecosystem metabolism analysis in China: A multi-scale integrated analysis of societal metabolism (MSIASM) approach[J]. Energy, 36(8): 4799-4808. doi: 10.1016/j.energy.2011.05.014
[13]	Giampietro M, Mayumi K.1997. A dynamic model of socioeconomic systems based on hierarchy theory and its application to sustainability[J]. Structural Change & Economic Dynamics, 8(4): 453-469.
[14]	Giampietro M, Mayumi K.2000. Multiple-scale integrated assessment of societal metabolism: Introducing the approach[J]. Population & Environment, 22(2): 109-153.
[15]	Giampietro M, Mayumi K, Munda G.2006. Integrated assessment and energy analysis: Quality assurance in multi-criteria analysis of sustainability[J]. Energy, 31(1): 59-86. doi: 10.1016/j.energy.2005.03.005
[16]	Greff K, Srivastava R K, Koutník J, et al.2017. LSTM: A search space odyssey[J]. IEEE Transactions on Neural Networks & Learning Systems, 28(10): 2222-2232.
[17]	Hadiye A S, Giampietro M, Aleu A L, et al.2009. Applications of the MuSIASEM approach to study changes in the metabolic pattern of Catalonia [R/OL]. Reports on Environmental Sciences. 2009-07-27 [2019-03-15]. .
[18]	Han W, Geng Y, Lu Y, et al.2018. Urban metabolism of megacities: A comparative analysis of Shanghai, Tokyo, London and Paris to inform low carbon and sustainable development pathways[J]. Energy, 155(15): 887-898. doi: 10.1016/j.energy.2018.05.073
[19]	Jiang Q, Tang C, Chen C, et al.2018. Stock price forecast based on LSTM neural network[C]// Xu J, Cooke F, Gen M, et al. Proceedings of the twelfth international conference on management science and engineering management. ICMSEM 2018. Cham: Springer: 393-408.
[20]	Kumar J, Goomer R, Singh A K.2018. Long short term memory recurrent neural network (LSTM-RNN) based workload forecasting model for cloud datacenters[J]. Procedia Computer Science, 125: 676-682. doi: 10.1016/j.procs.2017.12.087
[21]	Li Y, Wang X, Tian X, et al.2018. Understanding the mechanism of urban material metabolism with ecological network analysis: An experimental study of Wuxi, China[J]. Ecological Modelling, 367: 58-67 doi: 10.1016/j.ecolmodel.2017.09.007
[22]	Lu Y, Geng Y, Qian Y, et al.2016. Changes of human time and land use pattern in one mega city's urban metabolism: A multi-scale integrated analysis of Shanghai[J]. Journal of Cleaner Production, 133(1): 391-401. doi: 10.1016/j.jclepro.2016.05.174
[23]	Ma X, Tao Z, Wang Y, et al.2015. Long short-term memory neural network for traffic speed prediction using remote microwave sensor data[J]. Transportation Research Part C, 54: 187-197 doi: 10.1016/j.trc.2015.03.014
[24]	Martín J R, Eisenmenger N, Schandl H.2008. Different trajectories of exosomatic energy metabolism for Brazil, Chile and Venezuela: Using the MSIASM approach [R]. Working Papers. Barcelona, Spain: Department of Applied Economics, Universitat Autonoma of Barcelona.
[25]	Mayumi K.1991. Temporary emancipation from land: From the industrial revolution to the present time[J]. Ecological Economics, 4(1): 35-56. doi: 10.1016/0921-8009(91)90004-X
[26]	Munro D A, Holdgate M W.1991. Caring for the earth: A strategy for sustainable living[M]. IUCN, UNEP, WWF: 14-15.
[27]	Pincetl S, Bunje P, Holmes T.2012. An expanded urban metabolism method: Toward a systems approach for assessing urban energy processes and causes[J]. Landscape & Urban Planning, 107(3): 193-202.
[28]	Qi W, Deng X, Chu X, et al.2017. Emergy analysis on urban metabolism by counties in Beijing[J]. Physics & Chemistry of the Earth Parts A/B/C, 101: 157-165.
[29]	Ramos-Martin J, Giampietro M, Mayumi K.2007. On China's exosomatic energy metabolism: An application of multi-scale integrated analysis of societal metabolism (MSIASM)[J]. Ecological Economics, 63(1): 174-191. doi: 10.1016/j.ecolecon.2006.10.020
[30]	Scheidel A. 2013. From MuSIASEM theory to practice: Report and reflections from field research in Kampot province, Cambodia [R/OL]. Reports on Environmental Sciences. Institute of Environmental Science and Technology (ICTA). Barcelona, Spain: Autonomous University of Barcelona (UAB). .
[31]	Silva-Macher J C.2016. A metabolic profile of peru: An application of multi-scale integrated analysis of societal and ecosystem metabolism (MuSIASEM) to the mining sector's exosomatic energy flows[J]. Journal of Industrial Ecology, 20(5): 1072-1082. doi: 10.1111/jiec.12337
[32]	Song T, Cai J M, Chahine T, et al.2014. Modeling urban metabolism of Beijing City, China: With a coupled system dynamics: Emergy model[J]. Stochastic Environmental Research & Risk Assessment, 28(6): 1511-1524.
[33]	Tejedor-Flores N, Vicente-Galindo P, Galindo-Villardón P.2017. Sustainability multivariate analysis of the energy consumption of ecuador using MuSIASEM and BIPLOT approach[J]. Sustainability, 9(6): 984. doi: 10.3390/su9060984
[34]	Vafa-Arani H, Jahani S, Dashti H, et al.2014. A system dynamics modeling for urban air pollution: A case study of Tehran, Iran[J]. Transportation Research Part D Transport & Environment, 31(31): 21-36.
[35]	Wang X, Wu S, Li S.2017. Urban metabolism of three cities in Jing-Jin-Ji urban agglomeration, China: Using the MuSIASEM Approach[J]. Sustainability, 9(8): 1481-1502. doi: 10.3390/su9081481
[36]	Yuan X, Chen C, Lei X, et al.2018. Monthly runoff forecasting based on LSTM-ALO model[J]. Stochastic Environmental Research & Risk Assessment, 32(8): 1-14.
[37]	Yan Z, Li Y, Zheng H.2017. Ecological network analysis of energy metabolism in the Beijing-Tianjin-Hebei (Jing-Jin-Ji) urban agglomeration[J]. Ecological Modelling, 351: 51-62. doi: 10.1016/j.ecolmodel.2017.02.015
[38]	Yan Z, Zheng H, Yang Z, et al.2016. Urban energy flow processes in the Beijing-Tianjin-Hebei (Jing-Jin-Ji) urban agglomeration: Combining multi-regional input-output tables with ecological network analysis[J]. Journal of Cleaner Production, 114: 243-256. doi: 10.1016/j.jclepro.2015.06.093
[39]	Zhai M, Huang G, Liu L, et al.2018. Dynamic input-output analysis for energy metabolism system in the Province of Guangdong, China[J]. Journal of Cleaner Production, 196: 747-762. doi: 10.1016/j.jclepro.2018.06.084
[40]	Zhang J, Zhang Y.2011. Ecological network analysis of an urban energy metabolic system[J]. Stochastic Environmental Research & Risk Assessment, 25(5): 685-695.

年份	第一产业		第二产业		第三产业		生活部门			总体
年份	EMR₁/(MJ/h)	增速/%	EMR₂/(MJ/h)	增速/%	EMR₃/(MJ/h)	增速/%	EMR₄/(MJ/h)	增速/%		EMR_T/(MJ/h)	增速/%
2017	22.36	1.09	143.31	1.3	52.04	2.08	2.89		2.62	10.29	-4.94
2018	22.11	-1.1	145.26	1.36	52.8	1.46	2.99		3.16	10	-2.77
2019	22.08	-0.15	147.23	1.36	53.63	1.57	3.09		3.37	9.55	-4.5
2020	21.95	-0.61	149.23	1.36	54.47	1.56	3.21		4.02	9.16	-4.13
2021	21.86	-0.41	151.26	1.36	55.33	1.57	3.35		4.26	8.69	-5.08
2022	21.75	-0.51	153.32	1.36	56.2	1.59	3.51		4.94	8.21	-5.47

代谢率	情景分类	2018年	2019年	2020年	2021年	2022年
EMR_T	历史情景	10.00	9.55	9.16	8.69	8.21
	低增长率	9.10	8.76	8.22	7.74	7.18
	高增长率	10.90	10.33	10.07	9.61	9.23
EMR₁	历史情景	22.11	22.08	21.95	21.86	21.75
	低增长率	20.29	20.43	20.14	20.05	19.86
	高增长率	23.88	23.71	23.72	23.65	23.61
EMR₂	历史情景	145.26	147.23	149.23	151.26	153.32
	低增长率	131.48	133.26	135.06	136.89	138.74
	高增长率	159.05	161.22	163.43	165.67	167.94
EMR₃	历史情景	52.80	53.63	54.47	55.33	56.20
	低增长率	48.20	48.98	49.72	50.49	51.27
	高增长率	57.38	58.30	59.28	60.29	61.34
EMR₄	历史情景	2.99	3.09	3.21	3.35	3.51
	低增长率	2.73	2.86	2.91	3.02	3.12
	高增长率	3.26	3.33	3.54	3.71	3.97

代谢率	情景分类	2018年	2019年	2020年	2021年	2022年	代数平均值
EMR_T	低增长率	-9.02	-8.27	-10.22	-10.90	-12.57	-10.19
EMR_T	高增长率	9.02	8.14	9.95	10.56	12.31	10.00
EMR₁	低增长率	-8.27	-7.48	-8.22	-8.25	-8.69	-8.18
EMR₁	高增长率	8.01	7.36	8.09	8.19	8.58	8.05
EMR₂	低增长率	-9.49	-9.49	-9.50	-9.50	-9.51	-9.50
EMR₂	高增长率	9.50	9.50	9.51	9.52	9.53	9.51
EMR₃	低增长率	-8.72	-8.68	-8.71	-8.74	-8.78	-8.73
EMR₃	高增长率	8.67	8.71	8.84	8.98	9.14	8.87
EMR₄	低增长率	-8.70	-7.20	-9.45	-9.65	-11.29	-9.26
EMR₄	高增长率	9.20	7.85	10.27	10.82	13.09	10.25

代谢率	EMR₁	EMR₂	EMR₃	EMR₄
EMR₁	0.7977
EMR₂	0.9559	0.8948
EMR₃	0.9531	0.9607	0.8209
EMR₄	0.9202	0.9862	0.9397	0.8464

模型	EMR_T	EMR₁	EMR₂	EMR₃	EMR₄
KPCA-SVR	0.0616	0.1636	0.1048	0.4116	0.2650
KPCA-REG	0.0484	0.1879	0.1321	0.3079	6.0582
KPCA-LAS	0.0883	0.2021	0.1321	0.3081	0.4219
KPCA-GBR	0.0864	0.1133	0.2592	0.4692	0.3193
LSMT	0.0470	0.0476	0.0779	0.0208	0.0585

北京城市代谢预测研究——基于长短期记忆神经网络模型

Urban metabolism prediction of Beijing City based on long short-term memory neural network

RichHTML

PDF (PC)

赞

可视化

摘要/Abstract

引用本文

使用本文

图/表 10

参考文献 40

相关文章 15

Metrics

本文评价

推荐阅读 0

[1]	朱冠宇, 陈鹏. 基于共同犯罪的犯罪人地域关系网络的空间演化及其影响因素——以北京市诈骗案件为例[J]. 地理科学进展, 2020, 39(5): 792-803.
[2]	郭雅琦, 陈鹏. 基于时空临近重复效应的犯罪热点特征及成因分析——以北京市抢劫案件为例[J]. 地理科学进展, 2020, 39(5): 804-814.
[3]	胡庆芳, 曹士圯, 杨辉斌, 王银堂, 李伶杰, 王立辉. 汉江流域安康站日径流预测的LSTM模型初步研究[J]. 地理科学进展, 2020, 39(4): 636-642.
[4]	卢明华, 周悦颜, 刘汉初, 许欣. 北京企业对河北直接投资的时空动态特征及影响因素[J]. 地理科学进展, 2020, 39(3): 389-401.
[5]	李晓丽, 宋伟轩, 吴威, 马雨竹. 基于跟踪调查的北京城市空间感知研究——以中国科学院大学硕士研究生为例[J]. 地理科学进展, 2020, 39(2): 276-285.
[6]	王蓓, 王良, 刘艳华, 杨波, 黄晓春, 杨明. 基于手机信令数据的北京市职住空间分布格局及匹配特征[J]. 地理科学进展, 2020, 39(12): 2028-2042.
[7]	李春江, 马静, 柴彦威, 关美宝. 居住区环境与噪音污染对居民心理健康的影响——以北京为例[J]. 地理科学进展, 2019, 38(7): 1103-1110.
[8]	申犁帆, 张纯, 李赫, 王烨, 王子甲. 城市轨道交通通勤与职住平衡状况的关系研究——基于大数据方法的北京实证分析[J]. 地理科学进展, 2019, 38(6): 791-806.
[9]	黄幸, 刘玉亭. 消费端视角的中国大城市新建绅士化现象——以北京宣武门ZS小区为例[J]. 地理科学进展, 2019, 38(4): 577-587.
[10]	杨振山, 吴笛, 杨定. 迁居意愿、地方依赖和社区认同——北京中关村地区居住选择调查分析[J]. 地理科学进展, 2019, 38(3): 417-427.
[11]	张延吉, 朱春武, 秦波. 犯罪数量与危害的空间分布及建成环境影响——基于北京市刑事案件的纵向研究[J]. 地理科学进展, 2019, 38(12): 1876-1889.
[12]	丁凯熙, 张利平, 宋晓猛, 佘敦先, 夏军. 北京地区汛期降雨时空演变特征及城市化影响研究[J]. 地理科学进展, 2019, 38(12): 1917-1932.
[13]	杜超, 王姣娥, 刘斌全, 黄鼎曦. 城市道路与公共交通网络中心性对住宅租赁价格的影响研究——以北京市为例[J]. 地理科学进展, 2019, 38(12): 1831-1842.
[14]	文嫮, 张强国, 杜恒, 罗芬. 北京电影产业空间集聚与网络权力分布特征研究[J]. 地理科学进展, 2019, 38(11): 1747-1758.
[15]	申犁帆, 张纯, 李赫, 王烨. 大城市通勤方式与职住失衡的相互关系[J]. 地理科学进展, 2018, 37(9): 1277-1290.