Identificação do Comportamento de Motoristas: Uma Abordagem Baseada em Teoria da Informação

• Micael S. Santos UFAL
• Gean S. Santos UFAL
• Andre L. L. Aquino UFAL

DOI: https://doi.org/10.5753/sbcup.2024.2389

Resumo

Neste trabalho, propomos a identificação do comportamento do motorista com o uso do algoritmo Random Forest e Long Short-Term Memory (LSTM), baseado em medidas de teoria da informação, como Entropia de Shannon, Complexidade Estatística e Informação de Fisher. Os modelos LSTM e Random Forest foram aplicados em dados provenientes dos sensores acelerômetro e giroscópio em veículos. Tais dados foram rotulados como: slow, normal, e aggressive. Comparamos a metodologia padrão da literatura com a nossa proposta por meio das medidas de acurácia, área sob a curva ROC (AUC), e precisão. Seguindo a literatura obtivemos: com Random Forest 60 % de acurácia, 58 % de AUC, e 61 % de precisão; com LSTM 56 à 58 % de acurácia, 52 % de AUC, e 68 à 73 % de precisão. Seguindo nossa proposta obtivemos: com Random Forest 52 à 92 % de acurácia, 54 à 94 % de AUC, e 61 à 100 % de precisão; com LSTM 61 à 80 % de acurácia, 58 à 78 % de AUC, e 56 à 85 % de precisão.

Referências

Aquino, A. L., Cavalcante, T. S., Almeida, E. S., Frery, A. C., and Rosso, O. A. (2015). Characterization of vehicle behavior with information theory. The European Physical Journal B, 88(10):257.

Bandt, C. and Pompe, B. (2002). Permutation entropy: A natural complexity measure for time series. Phys. Rev. Lett., 88:174102.

Bernardi, M. L., Cimitile, M., Martinelli, F., and Mercaldo, F. (2018). Driver and path detection through time-series classification. Journal of Advanced Transportation, 2018:1758731.

Cojocaru, I. and Popescu, P.-S. (2022). Building a driving behaviour dataset. In Romanian Conference on Human-Computer Interaction.

Girma, A., Yan, X., and Homaifar, A. (2019). Driver identification based on vehicle telematics data using lstm-recurrent neural network. In 2019 IEEE 31st International Conference on Tools with Artificial Intelligence (ICTAI), pages 894–902.

Lamberti, P. W., Martin, M., Plastino, A., and Rosso, O. A. (2004). Intensive entropic non-triviality measure. Physica A: Statistical Mechanics and its Applications, 334(1-2):119–131.

López-Ruiz, R., Mancini, H. L., and Calbet, X. (1995). A statistical measure of complexity. Physics Letters A, 209(5-6):321–326.

Manderna, A. and Kumar, S. (2022). Effective long short-term memory based-driver identification in its. In 2022 International Conference on Inventive Computation Technologies (ICICT).

Martinelli, F., Mercaldo, F., Orlando, A., Nardone, V., Santone, A., and Kumar, A. (2020). Human behavior characterization for driving style recognition in vehicle system. Computers & Electrical Engineering, 83.

Mateos, D. M., Gómez-Ramírez, J., and Rosso, O. A. (2021). Using time causal quantifiers to characterize sleep stages. Chaos, Solitons & Fractals, 146:110798.

Miyajima, C., Nishiwaki, Y., Ozawa, K., Wakita, T., Itou, K., Takeda, K., and Itakura, F. (2007). Driver modeling based on driving behavior and its evaluation in driver identification. Proceedings of the IEEE, 95(2):427–437.

Nascimento, W. S. and Prudente, F. V. (2016). Study of shannon entropy in the context of quantum mechanics: An application to free and confined harmonic oscillator. In Quim. Nova, pages 757–764.

Olivares, F., Souza, L., Legnani, W., and Rosso, O. A. (2019). Informational time causal planes: A tool for chaotic map dynamic visualization. In Legnani, W. and Moschandreou, T. E., editors, Nonlinear Systems-Theoretical Aspects and Recent Applications, chapter 5. IntechOpen, Rijeka.

Park, K. H. and Kim, H. K. (2019). This car is mine!: Automobile theft countermeasure leveraging driver identification with generative adversarial networks. CoRR, abs/1911.09870.

Pessa, A. A. B. and Ribeiro, H. V. (2021). ordpy: A python package for data analysis with permutation entropy and ordinal network methods. Chaos: An Interdisciplinary Journal of Nonlinear Science, 31(6):063110.

Popescu, P.-S. and Ion, C. (2022). Driving behavior.

Santos, G. S. (2024). Use of information theory measures extracted from obd-ii interface data for driver identification. Mestrado em informática, Universidade Federal de Alagoas, Instituto de Computação, Maceió.

Schneegass, S., Pfleging, B., Broy, N., Heinrich, F., and Schmidt, A. (2013). A data set of real world driving to assess driver workload. In Proceedings of the 5th International Conference on Automotive User Interfaces and Interactive Vehicular Applications, AutomotiveUI ’13, page 150–157, New York, NY, USA. Association for Computing Machinery.

Singh, M. and Dubey, R. K. (2023). Deep learning model based co2 emissions prediction using vehicle telematics sensors data. IEEE Transactions on Intelligent Vehicles, 8(1):768–777.

Uvarov, K. and Ponomarev, A. (2021). Driver identification with obd-ii public data. In 2021 28th Conference of Open Innovations Association (FRUCT), pages 495–501.

Vaiti, T., Tisljaric, L., Erdelic, T., and Caric, T. (2022). Traffic emissions clustering using obd-ii dataset based on machine learning algorithms. Transportation Research Procedia, 64:364–371. International Scientific Conference “The Science and Development of Transport Znanost i razvitak prometa”.

Wu, J.-D. and Ye, S.-H. (2009). Driver identification using finger-vein patterns with radon transform and neural network. Expert Syst. Appl., 36:5793–5799.

Xun, Y., Sun, Y., and Liu, J. (2019). An experimental study towards driver identification for intelligent and connected vehicles. In ICC 2019 2019 IEEE International Conference on Communications (ICC).