조건부 적대적 생성모델 기반 교통수요 분석의 맞춤형 합성인구 생성기법 연구
Customized Population Synthesis for Transportation Demand Analysis Based on Conditional Generative Adversarial Network
0
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
교통수요 분석에 적용되고 있는 생성모델은 현실에 존재하는 복합적인 통행 의사결정을 모사하기위한 활동기반 모델에 활발히 도입되고 있다. 이는 분석 대상지 내에 현실에서 수집될 수 없는 총인구정보를 취득하기 위해 주어진 데이터를 활용하여 합성인구 생성 및 위치정보를 포함한 개인단위의통행 일정 생성까지 통계적으로 유의미한 기초데이터를 모사하는 것을 목표로 한다. 그러나 교통수요분석을 위해 활용되는 데이터는 미래 시점의 인구정보와 인구구조의 변화(i.e. 통행목적, 수단분담률) 등에 따른 통행행태 변화를 분석할 수 있는 조건에 대해 유연하게 반영할 수 있어야 한다. 따라서본 연구에서는 조건부 적대적 생성모델을 활용하여 맞춤형 합성인구를 생성할 수 있는 방법론을제안한다. 가구통행실태조사 자료의 인구정보를 토대로 생성된 데이터의 전반적인 정확성과 다양성은정밀도(precision)와 재현율(recall)로 측정하여 F1 score 기반의 종합적인 평가를 수행하였고, 특정조건부에 따라 생성된 데이터는 실제 데이터에서 추출한 인구정보와 비교하여 정확성을 판단하였다. 제안된 모델을 기반으로 다양한 활동기반 모형 분석에 응용될 것으로 기대된다.
Abstract
The generative models applied to traffic demand analysis are actively being introduced into activity-based models that simulate complex travel decision-making in reality. These models aim to statistically replicate foundational data, including synthetic population generation and individual activity schedules with location information, based on the data available to acquire comprehensive population information that cannot be collected. Nevertheless, the data used for traffic demand analysis must flexibly reflect the conditions that allow for the analysis of future population information and changes in travel behavior (e.g., travel purpose and mode share) due to shifts in the population structure. Therefore, this study proposes a methodology that uses Conditional Generative Adversarial Networks (CGAN) to generate customized synthetic populations. The overall accuracy and diversity of the data generated based on the population information from the Household Travel Survey (HTS) were measured using precision and recall, and a comprehensive evaluation was performed using the F1 score. The accuracy of the data generated under specific conditions was also compared with actual population data. The proposed model can be applied to various activity-based model analyses.
참고문헌 (15)
- Aemmer, Z., "Generative population synthesis for joint household and individual characteristics", Computers, Environment and Urban Systems, 96, pp.101852~ (2022)
- Badu-Marfo, G., "Composite travel generative adversarial networks for tabular and sequential population synthesis", IEEE Transactions on Intelligent Transportation Systems, 23(10), pp.17976~17985 (2022)
- Borysov, S. S., "How to generate micro-agents? A deep generative modeling approach to population synthesis", Transportation Research Part C : Emerging Technologies, 106, pp.73~97 (2019)
- Chapuis, K., "Generation of synthetic populations in social simulations: A review of methods and practices", Journal of Artificial Societies and Social Simulation, 25(2) (2022)
- Choupani, A. A., "Population synthesis using iterative proportional fitting(IPF) : A review and future research", Transportation Research Procedia, 17, pp.223~233 (2016)
- Farooq, B., "Simulation based population synthesis", Transportation Research Part B : Methodological, 58, pp.243~263 (2013)
- Goodfellow, I., "Generative adversarial nets", Advances in Neural Information Processing Systems, 27 (2014)
- Gulrajani, I., "Improved training of wasserstein gans", Advances in Neural Information Processing Systems, 30 (2017)
- Hörl, S., "Synthetic population and travel demand for Paris and Île-de-France based on open and publicly available data", Transportation Research Part C: Emerging Technologies, 130, pp.103291~ (2021)
- Kim, E. J., "A deep generative model for feasible and diverse population synthesis", Transportation Research Part C: Emerging Technologies, 148, pp.104053~ (2023)
- Kim, E. J., "Imputing qualitative attributes for trip chains extracted from smart card data using a conditional generative adversarial network", Transportation Research Part C: Emerging Technologies, 137, pp.103616~ (2022)
- Saadi, I., "Hidden markov model-based population synthesis", Transportation Research Part B : Methodological, 90, pp.1~21 (2016)
- Sun, L., "A hierarchical mixture modeling framework for population synthesis", Transportation Research Part B : Methodological, 114, pp.199~212 (2018)