Feasible and Novel Synthetic Population Generation with Tabular and Sequential Travel Attributes

2026-08-16

AI Summary

This study addresses the challenges of sampling zeros, structural infeasibility, and missing travel sequences in small-sample synthetic population generation by proposing a regularized two-stage generative framework. Integrating WGAN-GP with Transformer/LSTM-Attention architectures, the method incorporates regularization terms such as IGP to enhance generation quality and introduces a count-aware metric to evaluate the recovery of unseen combinations. Experimental results demonstrate that the model improves feasibility by 3.7%, novelty by 10%, and F1 score by 8.6%, achieving a sequence F1 of 90.6% with significant cross-stage consistency. These findings confirm the framework’s effectiveness in jointly generating high-quality synthetic populations and activity-travel chains, offering a robust solution for data-sparse scenarios in urban mobility modeling.

Abstract

Synthetic populations are critical inputs for activity-based travel demand models, yet generating realistic populations from limited survey data remains challenging. Small samples miss valid attribute combinations, known as sampling zeros, and generative models may also produce infeasible structural zeros. Moreover, realistic synthetic populations must capture both static socio-demographic attributes and sequential travel behaviour, such as trip chains. This paper proposes a regularized two-stage generative framework to address these challenges, where regularization refers to additional loss terms that guide the generator toward broader valid coverage and fewer infeasible samples. In Stage 1, a Wasserstein GAN with gradient penalty is augmented with three regularization terms, IGP, LDR, and CLAP, to improve feasibility, diversity, and novelty in tabular population synthesis. In Stage 2, Transformer and LSTM-Attention models generate sequential travel attributes, including departure time, trip purpose, and travel mode, conditioned on the synthesized tabular profiles. We also introduce novelty and count-aware metrics to evaluate whether valid unseen combinations are recovered and generated in realistic proportions. Results show that regularized models outperform the vanilla WGAN-GP across feasibility, diversity, and novelty. Regularization increases feasibility by 2.1 to 3.7 percentage points and novelty by 6.6 to 10.0 percentage points, improving sampling-zero recovery without sacrificing feasibility. The F1 score improves by 6.3 to 8.6 percentage points. For sequential attributes, LSTM-Attention best matches the trip-length distribution, while Transformer achieves higher overall sequential F1, 90.6\% versus 89.1\%. Cross-stage validation confirms strong consistency between generated mobility status and generated trip chains.

Problem

• Synthetic Population Generation
• Activity-based Travel Demand Models
• Sampling Zeros
• Structural Zeros
• Sequential Travel Attributes

Innovation

• Regularized WGAN-GP
• Two-stage Generative Framework
• Sequential Travel Attributes
• Novelty Metrics
• Synthetic Population Generation