Forecasting Multivariate Urban Data via Decomposition and Spatio-Temporal Graph Analysis

2025-05-28

AI Summary

To address the challenge of joint forecasting of multivariate urban indicators—such as electricity consumption, meteorological variables, carbon intensity, and air quality—characterized by complex, dynamic spatiotemporal dependencies, this paper proposes a novel method integrating time-series decomposition with dynamic graph neural networks. It models multivariate indicators as graph nodes, enhances interpretability via STL (Seasonal-Trend-Residual) decomposition, and introduces a dynamic graph construction mechanism coupled with a graph attention network (GAT) to capture heterogeneous cross-variable dependencies. A multi-step rolling prediction architecture enables end-to-end joint modeling. Evaluated on real-world urban datasets, the method achieves an average 18.7% reduction in MAE over strong baselines including Informer and Autoformer. It supports high-accuracy, 72-hour collaborative forecasting and has been deployed in a smart-city energy dispatch prototype system.

Technology Category

• Planning, Routing, and Scheduling: Optimization of Spatio-temporal Systems
• Machine Learning: Graph-based Machine Learning
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Systems and Infrastructure for Web, Mobile and WoT: Energy management for devices in mobile Web and WoT environments

Abstract

The forecasting of multivariate urban data presents a complex challenge due to the intricate dependencies between various urban metrics such as weather, air pollution, carbon intensity, and energy demand. This paper introduces a novel multivariate time-series forecasting model that utilizes advanced Graph Neural Networks (GNNs) to capture spatial dependencies among different time-series variables. The proposed model incorporates a decomposition-based preprocessing step, isolating trend, seasonal, and residual components to enhance the accuracy and interpretability of forecasts. By leveraging the dynamic capabilities of GNNs, the model effectively captures interdependencies and improves the forecasting performance. Extensive experiments on real-world datasets, including electricity usage, weather metrics, carbon intensity, and air pollution data, demonstrate the effectiveness of the proposed approach across various forecasting scenarios. The results highlight the potential of the model to optimize smart infrastructure systems, contributing to energy-efficient urban development and enhanced public well-being.

Problem

• Forecasting multivariate urban data with complex dependencies
• Capturing spatial dependencies using Graph Neural Networks
• Improving forecast accuracy via decomposition-based preprocessing

Innovation

• Decomposes time-series into trend, seasonal, residual components
• Uses Graph Neural Networks for spatial dependencies
• Improves forecasting via dynamic GNN interdependencies