Fusing Urban Structure and Semantics: A Conditional Diffusion Model for Cross-City OD Matrix Generation

2026-05-01

AI Summary

Existing methods struggle to generate realistic inter-city commuting origin-destination (OD) matrices due to insufficient modeling of the high heterogeneity arising from individual intentions, geographic constraints, and social dynamics. This work represents a city as an attributed graph, where nodes correspond to regions enriched with population and POI features, and edges denote commuting flows. We propose a structure-enhanced conditional diffusion model that, for the first time in OD generation, jointly integrates regional semantic attributes with explicit spatial constraints—namely adjacency and distance matrices. The model employs a graph Transformer to capture node interactions and incorporates adjacency-guided attention alongside distance-aware diffusion conditioning to ensure behavioral plausibility and geographic consistency. Evaluated on real-world multi-city data from the United States, our approach reduces RMSE by 7.38% compared to the state-of-the-art WEDAN model and demonstrates strong robustness across heterogeneous urban structures.

Abstract

Accurate modeling of commuting flows is important for urban governance, traffic planning, and resource allocation. However, the combined influence of individual intentions, geographic constraints, and social dynamics leads to considerable heterogeneity in commuting patterns, making it difficult to develop generation models that generalize across cities. To address this issue, we propose SEDAN, a Structure-Enhanced Diffusion model conditioned on Attributed Nodes for generalizable OD matrix generation. SEDAN models a city as an attributed graph. Each region is treated as a node with demographic and point-of-interest features, and commuting flows are modeled as weighted edges. Adjacency and distance matrices are incorporated to characterize spatial structure. Based on this representation, we design a fusion mechanism within SEDAN to jointly model semantic information and spatial information. Regional semantic attributes are used to model latent travel demand through graph-transformer-based node interactions, while spatial structure is injected into the generation process as explicit constraints. The adjacency matrix guides attention weights to strengthen interactions between neighboring regions. Meanwhile, the distance matrix serves as a diffusion condition to capture spatial proximity and travel impedance. The fusion of urban semantics and spatial constraints enables SEDAN to generate OD matrices that are both behaviorally plausible and geographically coherent. Experiments on real-world OD datasets from U.S. cities show that SEDAN achieves a 7.38\% improvement in RMSE over the state-of-the-art baseline, WEDAN. It also remains robust across heterogeneous urban scenarios and varying structural patterns. Our work provides an effective and generalizable solution for commuting OD matrix generation. The code is available at https://anonymous.4open.science/r/SEDAN.

Problem

• OD matrix generation
• cross-city generalization
• urban heterogeneity
• commuting flow modeling
• spatial structure

Innovation

• conditional diffusion model
• OD matrix generation
• attributed graph
• spatial-semantic fusion
• graph transformer