Copula-based scenario generation for urban traffic models
作者：
Highlights：
• Statistically-sound what-if scenario generation for urban traffic models.
• Data-driven generation through copula models trained on available data.
• The method can be applied to any traffic model and arbitrary sets of input parameters.
• Joint dependence of generated inputs is preserved, beyond simple correlation analysis.
• Simulation results are provided for both a large-scale and a microscopic model.
摘要
•Statistically-sound what-if scenario generation for urban traffic models.•Data-driven generation through copula models trained on available data.•The method can be applied to any traffic model and arbitrary sets of input parameters.•Joint dependence of generated inputs is preserved, beyond simple correlation analysis.•Simulation results are provided for both a large-scale and a microscopic model.
论文关键词：Generative models,Urban traffic models,Copulas,Scenario generation
论文评审过程：Received 23 April 2022, Revised 18 July 2022, Accepted 1 August 2022, Available online 6 August 2022, Version of Record 15 August 2022.
论文官网地址：https://doi.org/10.1016/j.eswa.2022.118389
