X-ST-SVDD: Arquitetura Explicável para Detecção de Anomalias Espaço-Temporais em Redes de Micromobilidade Urbana
Resumo
Infraestruturas de mobilidade urbana, como compartilhamento de bicicletas, exigem monitoramento em tempo real para disrupções ambientais. Esta modelagem é dificultada pela topologia não-euclidiana das redes de transporte e escassez de anomalias rotuladas. Este trabalho apresenta a X-ST-SVDD, uma arquitetura explicável de detecção de anomalias espaço-temporais para micromobilidade. A solução combina Redes Neurais em Grafos Espaço-Temporais (ST-GNN) e o algoritmo Deep SVDD para aprender padrões normais de operação e detectar desvios de forma não supervisionada. Mecanismos de explicabilidade baseados em gradientes são integrados para quantificar a contribuição causal de cada variável. A avaliação utilizou uma rede simulada de 120 estações durante 100 dias, combinando variáveis meteorológicas e de demanda. Sob um cenário de disrupção climática severa, os resultados demonstram que a arquitetura distingue com precisão estados normais de anômalos, amplificando o escore de anomalia em mais de 740 vezes e fornecendo diagnósticos causais interpretáveis.
Referências
Adnan, M. H., Zukarnain, Z. A., and Subramaniam, S. K. (2025). Path planning for multi-UAV-assisted mobile edge computing framework using reinforcement learning in urban environment disaster scenarios. Ad Hoc Networks, 178:103928.
Cai, J., Zhou, Y., Wang, Y., Xiao, Y., and Cai, R. (2026). Multi-scale spatiotemporal analysis of shared bike-metro interactions in Shanghai: Implications for sustainable urban mobility. Research in Transportation Business & Management, 66:101627.
Chen, K. and Wang, X. (2026). The impact of intelligent computing centers construction on urban new-quality productivity: An empirical study based on 287 prefecture-level cities in China. Socio-Economic Planning Sciences, 105:102476.
Chen, T., Sun, J., Zhang, Z., Xiao, Z., Zheng, L., Chai, H., and Lin, B. (2025). High-performance computing in urban flood modeling: A study on spatial partitioning techniques and parallel performance. Journal of Hydrology, 649:132474.
Chen, Z., Zhang, X., Li, S., Wu, T., Qiao, R., Zhao, J., Zhou, S., and Wu, Z. (2026). Interactive effects of physical environment and socio-economic factor on urban mobility resilience: The case of bike-sharing in Shanghai. Cities, 169:106550.
Deng, Z., Xu, Y., and Hong, T. (2026). Quantum computing approach for building surface sunlit in urban-scale energy modeling. Energy & Buildings, 353:116898.
Fu, L., Xu, Y., and Gao, S. (2025). A multi-agent deep distribution approximation strategy optimization algorithm with multi-threaded parallel computing mechanism suitable for large-scale and complex urban road networks. Engineering Applications of Artificial Intelligence, 155:110999.
Gao, H., Dong, Y., Zheng, R., Zheng, L., and Li, Y. (2026a). Registration generation amalgamation network for few-shot anomaly detection. Neurocomputing.
Gao, W., Yu, Z., Wang, T., Wang, L., Cui, H., Guo, B., and Xiong, H. (2025). GNN-based deep reinforcement learning for computation task scheduling in autonomous multi-robot systems. Journal of Systems Architecture, 168:103534.
Gao, Z., Wei, X., Li, H., Fan, Y., and Zhang, X. (2026b). Optimized acoustic computed tomography for monitoring urban road tunnel fire-induced ceiling temperatures using simulated annealing and SART. Mechanical Systems and Signal Processing, 247:113930.
Jia, J., He, S., Zhang, H., and Xiao, Y. (2025). Unveiling pandemic-driven mobility shifts: A S-GTWR analysis of bike-sharing and taxi systems in Washington, D.C. Sustainable Cities and Society, 132:106802.
Kang, Y. (2024). Visualization analysis of urban planning assistant decision network 3D system based on intelligent computing. Heliyon, 10:e31321.
Khattak, S. A., Liu, Y.-K., Liu, Y.-K., Gao, J.-R., Shi, Z.-X., and Liu, J. (2026). A deep learning-based prognostic approach for predicting PWR degradation and remaining useful life using GNN-PTC-LSTM. Annals of Nuclear Energy, 230:112172.
Kim, J., Suh, B., Park, Y., Kim, K.-I., Ullah, S., and Bhargava, A. (2026). Recent trends in artificial intelligence for anomaly detection and their applications in low power wireless networks. Pervasive and Mobile Computing, 119:102214.
Lei, B., Li, N., Guo, Y., Wang, Z., Wei, J., and Chen, R. (2024). Rapid data collection and processing in dense urban edge computing networks with drone assistance. Physical Communication, 66:102462.
Li, W., Song, X., and Tu, Y. (2025a). GraphDRL: GNN-based deep reinforcement learning for interactive recommendation with sparse data. Expert Systems With Applications, 273:126832.
Li, Y., Xiong, B., Wang, C., Sun, T., Xiong, Q., and Kong, Q. (2025b). Automatic, batching and remote detection of urban building inclination information with edge computing of LIDAR point clouds on a UAV. Journal of Building Engineering, 109:112981.
Mahajan, P. and Kaul, A. (2025). Graph-enhanced deep learning for ECG arrhythmia detection: An integration of CNN-GNN-BILSTM approach. Medical Engineering and Physics, 145:104418.
Meneguette, R., De Grande, R., Ueyama, J., Filho, G. P. R., and Madeira, E. (2021). Vehicular edge computing: Architecture, resource management, security, and challenges. ACM Computing Surveys (CSUR), 55(1):1–46.
Mnasri, Z. and Bouwmans, T. (2026). A novel method for anomaly detection using graph signal embedding: Application to anomalous sound detection. Expert Systems with Applications, page 132278.
Rocha Filho, G. P., Meneguette, R. I., Neto, J. R. T., Valejo, A., Weigang, L., Ueyama, J., Pessin, G., and Villas, L. A. (2020). Enhancing intelligence in traffic management systems to aid in vehicle traffic congestion problems in smart cities. Ad Hoc Networks, 107:102265.
Sanudo, E., García-Feal, O., Hagen, L., Cea, L., Puertas, J., Montalvo, C., Alvarado-Vicencio, R., and Hofmann, J. (2025). IberSWMM+: A high-performance computing solver for 2D-1D pluvial flood modelling in urban environments. Journal of Hydrology, 651:132603.
Sruthi, K., Sreekumar, A., and Balakrishnan, K. (2026). Text meets topology: A dynamic graph learning method with Dy-GNNs and LLMs for tracking text attributed community evolution in complex networks. Chaos, Solitons and Fractals, 208:118121.
Xu, X., Wang, J., Poslad, S., Rui, X., Zhang, G., and Fan, Y. (2026). Exploring urban human mobility patterns and the effects of air pollution and weather conditions from a dockless bike-sharing lens. Journal of Cleaner Production, 554:148155.
Yin, A., Chen, X., He, H., Morris, A., Yuan, Q., Ma, X., and Yang, Z. (2025). Shared e-bikes demand in urban mobility: Temporal heterogeneity, driving factors, and strategic implications. Travel Behaviour and Society, 41:101075.
Zhang, S., Chen, L., Xu, L., and Wang, Z. (2025). GeoBEM: A geospatial computing empowered framework for urban-scale building energy modeling. Sustainable Cities and Society, 121:106203.
Zhong, Y., Guo, B., Sun, Z., Deng, L., Xu, X., and Dian, S. (2026). MSDCIR-AD: Unsupervised anomaly detection via multi-criteria semantic distances and constrained image reconstruction. Computers in Industry, 178:104467.
Zhu, H., Liu, X., Zhang, Y., and Zhang, L. (2026). A new reconstruction-based method for multivariate time series anomaly detection with diffusion models. Engineering Applications of Artificial Intelligence, 175:114656.
Zou, X., Yan, Y., Hao, X., Hu, Y., Wen, H., Liu, E., Zhang, J., Li, Y., Li, T., Zheng, Y., and Liang, Y. (2025). Deep learning for cross-domain data fusion in urban computing: Taxonomy, advances, and outlook. Information Fusion, 113:102606.
Cai, J., Zhou, Y., Wang, Y., Xiao, Y., and Cai, R. (2026). Multi-scale spatiotemporal analysis of shared bike-metro interactions in Shanghai: Implications for sustainable urban mobility. Research in Transportation Business & Management, 66:101627.
Chen, K. and Wang, X. (2026). The impact of intelligent computing centers construction on urban new-quality productivity: An empirical study based on 287 prefecture-level cities in China. Socio-Economic Planning Sciences, 105:102476.
Chen, T., Sun, J., Zhang, Z., Xiao, Z., Zheng, L., Chai, H., and Lin, B. (2025). High-performance computing in urban flood modeling: A study on spatial partitioning techniques and parallel performance. Journal of Hydrology, 649:132474.
Chen, Z., Zhang, X., Li, S., Wu, T., Qiao, R., Zhao, J., Zhou, S., and Wu, Z. (2026). Interactive effects of physical environment and socio-economic factor on urban mobility resilience: The case of bike-sharing in Shanghai. Cities, 169:106550.
Deng, Z., Xu, Y., and Hong, T. (2026). Quantum computing approach for building surface sunlit in urban-scale energy modeling. Energy & Buildings, 353:116898.
Fu, L., Xu, Y., and Gao, S. (2025). A multi-agent deep distribution approximation strategy optimization algorithm with multi-threaded parallel computing mechanism suitable for large-scale and complex urban road networks. Engineering Applications of Artificial Intelligence, 155:110999.
Gao, H., Dong, Y., Zheng, R., Zheng, L., and Li, Y. (2026a). Registration generation amalgamation network for few-shot anomaly detection. Neurocomputing.
Gao, W., Yu, Z., Wang, T., Wang, L., Cui, H., Guo, B., and Xiong, H. (2025). GNN-based deep reinforcement learning for computation task scheduling in autonomous multi-robot systems. Journal of Systems Architecture, 168:103534.
Gao, Z., Wei, X., Li, H., Fan, Y., and Zhang, X. (2026b). Optimized acoustic computed tomography for monitoring urban road tunnel fire-induced ceiling temperatures using simulated annealing and SART. Mechanical Systems and Signal Processing, 247:113930.
Jia, J., He, S., Zhang, H., and Xiao, Y. (2025). Unveiling pandemic-driven mobility shifts: A S-GTWR analysis of bike-sharing and taxi systems in Washington, D.C. Sustainable Cities and Society, 132:106802.
Kang, Y. (2024). Visualization analysis of urban planning assistant decision network 3D system based on intelligent computing. Heliyon, 10:e31321.
Khattak, S. A., Liu, Y.-K., Liu, Y.-K., Gao, J.-R., Shi, Z.-X., and Liu, J. (2026). A deep learning-based prognostic approach for predicting PWR degradation and remaining useful life using GNN-PTC-LSTM. Annals of Nuclear Energy, 230:112172.
Kim, J., Suh, B., Park, Y., Kim, K.-I., Ullah, S., and Bhargava, A. (2026). Recent trends in artificial intelligence for anomaly detection and their applications in low power wireless networks. Pervasive and Mobile Computing, 119:102214.
Lei, B., Li, N., Guo, Y., Wang, Z., Wei, J., and Chen, R. (2024). Rapid data collection and processing in dense urban edge computing networks with drone assistance. Physical Communication, 66:102462.
Li, W., Song, X., and Tu, Y. (2025a). GraphDRL: GNN-based deep reinforcement learning for interactive recommendation with sparse data. Expert Systems With Applications, 273:126832.
Li, Y., Xiong, B., Wang, C., Sun, T., Xiong, Q., and Kong, Q. (2025b). Automatic, batching and remote detection of urban building inclination information with edge computing of LIDAR point clouds on a UAV. Journal of Building Engineering, 109:112981.
Mahajan, P. and Kaul, A. (2025). Graph-enhanced deep learning for ECG arrhythmia detection: An integration of CNN-GNN-BILSTM approach. Medical Engineering and Physics, 145:104418.
Meneguette, R., De Grande, R., Ueyama, J., Filho, G. P. R., and Madeira, E. (2021). Vehicular edge computing: Architecture, resource management, security, and challenges. ACM Computing Surveys (CSUR), 55(1):1–46.
Mnasri, Z. and Bouwmans, T. (2026). A novel method for anomaly detection using graph signal embedding: Application to anomalous sound detection. Expert Systems with Applications, page 132278.
Rocha Filho, G. P., Meneguette, R. I., Neto, J. R. T., Valejo, A., Weigang, L., Ueyama, J., Pessin, G., and Villas, L. A. (2020). Enhancing intelligence in traffic management systems to aid in vehicle traffic congestion problems in smart cities. Ad Hoc Networks, 107:102265.
Sanudo, E., García-Feal, O., Hagen, L., Cea, L., Puertas, J., Montalvo, C., Alvarado-Vicencio, R., and Hofmann, J. (2025). IberSWMM+: A high-performance computing solver for 2D-1D pluvial flood modelling in urban environments. Journal of Hydrology, 651:132603.
Sruthi, K., Sreekumar, A., and Balakrishnan, K. (2026). Text meets topology: A dynamic graph learning method with Dy-GNNs and LLMs for tracking text attributed community evolution in complex networks. Chaos, Solitons and Fractals, 208:118121.
Xu, X., Wang, J., Poslad, S., Rui, X., Zhang, G., and Fan, Y. (2026). Exploring urban human mobility patterns and the effects of air pollution and weather conditions from a dockless bike-sharing lens. Journal of Cleaner Production, 554:148155.
Yin, A., Chen, X., He, H., Morris, A., Yuan, Q., Ma, X., and Yang, Z. (2025). Shared e-bikes demand in urban mobility: Temporal heterogeneity, driving factors, and strategic implications. Travel Behaviour and Society, 41:101075.
Zhang, S., Chen, L., Xu, L., and Wang, Z. (2025). GeoBEM: A geospatial computing empowered framework for urban-scale building energy modeling. Sustainable Cities and Society, 121:106203.
Zhong, Y., Guo, B., Sun, Z., Deng, L., Xu, X., and Dian, S. (2026). MSDCIR-AD: Unsupervised anomaly detection via multi-criteria semantic distances and constrained image reconstruction. Computers in Industry, 178:104467.
Zhu, H., Liu, X., Zhang, Y., and Zhang, L. (2026). A new reconstruction-based method for multivariate time series anomaly detection with diffusion models. Engineering Applications of Artificial Intelligence, 175:114656.
Zou, X., Yan, Y., Hao, X., Hu, Y., Wen, H., Liu, E., Zhang, J., Li, Y., Li, T., Zheng, Y., and Liang, Y. (2025). Deep learning for cross-domain data fusion in urban computing: Taxonomy, advances, and outlook. Information Fusion, 113:102606.
Publicado
25/05/2026
Como Citar
CARDOSO, Diego Dias; ZAGO, Cassiano Darif; IMMICH, Roger; ROCHA FILHO, Geraldo Pereira.
X-ST-SVDD: Arquitetura Explicável para Detecção de Anomalias Espaço-Temporais em Redes de Micromobilidade Urbana. In: WORKSHOP DE COMPUTAÇÃO URBANA (COURB), 10. , 2026, Praia do Forte/BA.
Anais [...].
Porto Alegre: Sociedade Brasileira de Computação,
2026
.
p. 378-391.
ISSN 2595-2706.
DOI: https://doi.org/10.5753/courb.2026.24122.
