基于遗传算法的人工人口生成与应用研究

doi:10.16182/j.issn1004731x.joss.22-0525

摘要/Abstract

摘要：

高精度微观人口数据是疾病传播、交通出行、应急事件等仿真系统的关键基础数据之一，现实中多采用计算机生成人工人口进行模拟。出于计算效率和生成步骤规范化考虑，目前人工人口合成多采用迭代比例拟合法，但是它对基础数据要求严格，存在零单元和数据代表性偏差问题且无法同时保证个体和家庭层面的拟合。为克服这一不足，提出使用模拟退火算法生成初始解和精英选择策略的改进遗传算法合成人工人口，探究遗传算法的合成步骤与参数设置，并以生成的成都市人工人口为基础，进行生物气溶胶扩散仿真实验。实验表明，采用遗传算法合成人工人口的精度优于传统方法，对基础数据要求低，这一数据集可为仿真研究提供精细尺度的人口多属性信息。

关键词: 人工人口, 遗传算法, 模拟退火算法, 精英选择, 气溶胶扩散

Abstract:

High-precision micro-population data are one of the key basic data for simulation systems such as disease spread, traffic travel, and emergency events. In reality, computer-generated artificial populations are often used for simulation. Due to computational efficiency and standardization of generation steps, the iterative proportional fitting method is currently used for artificial population synthesis. However, it has strict requirements on basic data and faces zero-unit and data representational deviation problems, and it fails to guarantee the fitting at the individual and family levels at the same time. In order to overcome this deficiency, an improved genetic algorithm using a simulated annealing algorithm to generate an initial solution and elitist selection strategy is proposed to synthesize artificial populations and explore the synthetic steps and parameter setting of the genetic algorithm. Based on the generated artificial population in Chengdu, a simulation experiment of biological aerosol diffusion is carried out. Experiments show that the accuracy of artificial populations generated by the genetic algorithm is better than that generated by traditional methods and has low requirements for basic data. This data set can provide fine-scale multi-attribute population information for simulation research.

Key words: artificial population, genetic algorithm, simulated annealing algorithm, elitist selection, aerosol diffusion

中图分类号:

TP391.9

张红历,邓井双 . 基于遗传算法的人工人口生成与应用研究[J]. 系统仿真学报, 2023, 35(9): 1965-1974.

Zhang Hongli,Deng Jingshuang . Research on Artificial Population Generation and Application Based on Genetic Algorithm[J]. Journal of System Simulation, 2023, 35(9): 1965-1974.

图/表 14

表1

表2

表3

表4

表5

表6

表7

表8

表9

表10

表11

图1

图2

图3

参考文献 23

1	葛渊峥, 宋智超, 孟荣清. 人工社会大规模人工人口生成方法综述[J]. 系统仿真学报, 2019, 31(10): 1951-1959.
	Ge Yuanzheng, Song Zhichao, Meng Rongqing. The Method Summary of Generating Large-scale Artificial Population in an Artificial Society[J]. Journal of System Simulation, 2019, 31(10): 1951-1959.
2	刘列, 许晴, 祖正虎, 等. 针对多代户家庭的人工人口生成方法研究[J]. 生物技术通讯, 2016, 27 (2): 237-243.
	Liu Lie, Xu Qing, Zu Zhenghu, et al. A Generation Method for Synthetic Populations of Multi-generational Households[J]. Letters in Biotechnology, 2016, 27(2): 237-243.
3	Boyam Fabrice Yaméogo, Gastineau P, Hankach P, et al. Comparing Methods for Generating a Two-layered Synthetic Population[J]. Transportation Research Record, 2021, 2675(1): 136-147.
4	Deming W E, Stephan F F. On a Least Squares Adjustment of a Sampled Frequency Table When the Expected Marginal Totals Are Known[J]. The Annals of Mathematical Statistics, 1940, 11(4): 427-444.
5	Fienberg S E. An Iterative Procedure for Estimation in Contingency Tables[J]. The Annals of Mathematical Statistics, 1970, 41(3): 907-917.
6	Beckman R J, Baggerly K A, McKay M D. Creating Synthetic Baseline Populations[J]. Transportation Research Part A: Policy and Practice, 1996, 30(6): 415-429.
7	Arentze T, Timmermans H, Hofman F. Creating Synthetic Household Populations: Problems and Approach[J]. Transportation Research Record, 2007, 2014(1): 85-91.
8	Zhu Yi, Ferreira J Jr. Synthetic Population Generation at Disaggregated Spatial Scales for Land Use and Transportation Microsimulation[J]. Transportation Research Record, 2014, 2429(1): 168-177.
9	Wheaton W D, Cajka J C, Chasteen B M, et al. Synthesized Population Databases: A US Geospatial Database for Agent-based Models[J]. Methods Report(RTI Press), 2009(10): 905.
10	Ye Xin, Konduri K C, Pendyala R M, et al. Methodology to Match Distributions of Both Household and Person Attributes in Generation of Synthetic Populations[C]// Transportation Research Board 88th Annual Meeting. Washington DC, United States: The National Academies of Sciences, Engineering, and Medicine, 2009: 01128672.
11	Müller K A. Generalized Approach to Population Synthesis[D]. Zurich, Switzerland: Swiss Federal Institute of Technology in Zurich, 2017.
12	Rich J. Large-scale Spatial Population Synthesis for Denmark[J]. European Transport Research Review, 2018, 10(2): 63.
13	Lee D H, Fu Yingfei. Cross-entropy Optimization Model for Population Synthesis in Activity-based Microsimulation Models[J]. Transportation Research Record, 2011, 2255(1): 20-27.
14	Chapuis K, Taillandier P, Drogoul A. Generation of Synthetic Populations in Social Simulations: A Review of Methods and Practices[J]. Journal of Artificial Societies and Social Simulation, 2022, 25(2): 6.
15	Templ M, Meindl B, Kowarik A, et al. Simulation of Synthetic Complex Data: The R Package SimPop[J]. Journal of Statistical Software, 2017, 79(10): 1-38.
16	Williamson P, Birkin M, Rees P H. The Estimation of Population Microdata by Using Data from Small Area Statistics and Samples of Anonymised Records[J]. Environment and Planning A: Economy and Space, 1998, 30(5): 785-816.
17	Kim J, Lee S. A Reproducibility Analysis of Synthetic Population Generation[J]. Transportation Research Procedia, 2015, 6: 50-63.
18	Voas D, Williamson P. An Evaluation of the Combinatorial Optimisation Approach to the Creation of Synthetic Microdata[J]. International Journal of Population Geography, 2000, 6(5): 349-366.
19	Ryan J, Maoh H, Kanaroglou P. Population Synthesis: Comparing the Major Techniques Using a Small, Complete Population of Firms[J]. Geographical Analysis, 2009, 41(2): 181-203.
20	Garrido S, Borysov S S, Pereira F C, et al. Prediction of Rare Feature Combinations in Population Synthesis: Application of Deep Generative Modelling[J]. Transportation Research Part C: Emerging Technologies, 2020, 120: 102787.
21	龙瀛, 沈振江, 毛其智. 城市系统微观模拟中的个体数据获取新方法[J]. 地理学报, 2011, 66(3): 416-426.
	Long Ying, Shen Zhenjiang, Mao Qizhi. Retrieving Individual Attributes From Aggregate Dataset for Urban Micro-simulation: A Preliminary Exploration[J]. Acta Geographica Sinica, 2011, 66(3): 416-426.
22	葛渊峥. 基于agent的人工社会框架设计与生成方法[D]. 长沙: 国防科学技术大学, 2014.
	Ge Yuanzheng. An Agent-based Framework and Generation of An Artificial Society[D]. Changsha: National University of Defense Technology, 2014.
23	邱晓刚, 陈彬, 张鹏. 面向应急管理的人工社会构建与计算实验[M]. 北京: 科学出版社, 2017.

类别	变量名	说明
个体属性	Age	年龄，取值范围为0~95及以上；可根据年龄，进行成人(adult)与未成年人(children)区分
个体属性	Sex	性别，1-男性、2-女性
家庭属性	NP	家庭总人口
	Female	家庭中女性总人数
	Male	家庭中男性总人数

编号	NP	Adult	Children
1	3	2	1
2	2	1	1
3	4	2	2
4	4	3	1
5	1	1	0

类型	家庭户(户)					人口(人)
类型	1	2	3	4	5及5+	成年人	未成年人
合计	3					7
数量	1	1	0	1	0	4	3

约束条件	Estimated(E)	Observed(O)	Error \|(E)-(O)\|
1	1	1	0
2	0	1	1
3	1	0	1
4	1	1	0
5+	0	0	0
成年人	6	4	2
未成年人	2	3	1

约束条件	Estimated(E)	Observed(O)	Error \|(E)-(O)\|
1	1	1	0
2	0	1	1
3	0	0	0
4	2	1	1
5+	0	0	0
成年人	6	4	2
未成年人	3	3	0