Journal of Auto-vehicle Safety Association (자동차안전학회지)

Korean Auto-vehicle Safety Association (한국자동차안전학회)
권호 표지
DOI QR코드

DOI QR Code

Visualization of Vehicle Driving Patterns Using the LSTM-Autoencoder Algorithm with Attention Applied

Attention이 적용된 LSTM-Autoencoder 알고리즘을 사용한 차량 주행 패턴 시각화

  • Received : 2024.01.11
  • Accepted : 2024.08.31
  • Published : 2024.09.30
Download PDF
⟨ Previous Next ⟩

Abstract

Recent advancements in vehicle driving data analysis have attracted significant attention as a key research area aimed at optimizing driving behavior and enhancing vehicle performance. This study aims to collect operational data from drivers, including steering angle, longitudinal acceleration, brake usage, and wheel speed, and to cluster driving segments based on their characteristics using deep learning techniques. To achieve this, we employed a methodology that integrates the LSTM-Autoencoder model, which captures essential features of time-series data through compression and reconstruction, with an Attention mechanism. The driving segments were clustered and visually represented with distinct colors for clarity. Experimental results from the proposed approach demonstrated an over 96% accuracy in aligning input and output values, facilitating the clustering of driving segments based on their distinctive features. These findings provide valuable insights for improving driver behavior and optimizing vehicle assistance systems, potentially contributing to significant advancements in this field of research.

Keywords

Acknowledgement

이 논문은 정부(과학기술정보통신부)의 제원으로 한국연구재단의 지원을 받아 수행된 연구임(No.2022R1G1A10070561230382068210102).

References

  1. Xu, H., Gao, Y., Yu, F., and Darrell, T., 2017, End-to-end learning of driving models from large-scale video datasets. In Proceedings of the IEEE conference on computer vision and pattern recognition (pp. 2174~2182).
  2. Alkinani, M. H., Khan, W. Z., and Arshad, Q., 2020, "Detecting human driver inattentive and aggressive driving behavior using deep learning: Recent advances, requirements and open challenges," Ieee Access, 8, 105008~105030.
  3. Moujahid, A., Tantaoui, M. E., Hina, M. D., Soukane, A., Ortalda, A., ElKhadimi, A., and Ramdane-Cherif, A., 2018, "Machine learning techniques in ADAS: A review," International Conference on Advances in Computing and Communication Engineering (ICACCE), pp. 235~242, IEEE.
  4. Wan, F., Guo, G., Zhang, C., Guo, Q., and Liu, J. 2019, "Outlier detection for monitoring data using stacked autoencoder," IEEE Access, 7, 173827~173837.
  5. Bao, W., Yue, J., and Rao, Y., A deep learning framework for financial time series using stacked autoencoders and long-short term memory, PLoS ONE 2017, 12, e0180944.
  6. Gensler, A., Henze, J., Sick, B., and Raabe, N., Deep Learning for solar power forecasting-An approach using AutoEncoder and LSTM Neural Networks, In Proceedings of the 2016 IEEE International Conference on Systems, Man, and Cybernetics (SMC), Budapest, Hungary, 9-12 October 2016, pp. 2858~2865.
  7. Wei, Wangyang, Honghai Wu, and Huadong Ma, 2019, "An autoencoder and LSTM-based traffic flow prediction method," Sensors, 19(13), 2946.
  8. Schmidhuber, J., Deep Learning in neural networks: An overview, Neural Netw. 2015, 61, 85~117. doi:10.1016/j.neunet.2014.09.003.
  9. A. Vaswani, N. Shazeer, N. Parmar, J. Uszkoreit, L. Jones, A. N. Gomez, L. Kaiser, and I. Polosukhin, 2017, "Attention is all you need," in Advances in neural information processing systems, pp. 5998~6008.
  10. Kim, H. Y., Choi, S. W., and Huh, K. S., 2020, "Probabilistic vehicle trajectory prediction considering inter-vehicle interaction based on multi-head attention architecture," Transactions of the Korean Society of Automotive Engineers, 28(9), 645~652.