EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Research on Optimization of Intelligent Driving Path Planning Algorithm Based on Deep Learning

Dachuang Qian, Qingshou Lu

Data and Metadata · 13.04.2026 · 5 · 806

DOI: 10.56294/dm2026806 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Data and Metadata, online erschienen am 13.04.2026. Verfasst von Dachuang Qian und 1 weiteren Autor:innen. Die Arbeit nennt 16 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Introduction

With the rapid development of intelligent driving technology, accurate, efficient, and robust path planning has become essential for autonomous vehicles, as traditional algorithms often exhibit limited adaptability and suboptimal performance in complex, dynamic, and uncertain road environments.

Objective

To address these challenges, this research proposes an optimized intelligent driving path planning algorithm based on DL, which integrates intelligent driving path planning contains 4,000 rows and 16 columns from Kaggle to improve both path quality and planning efficiency.

Method

Environmental perception through a multi-modal approach to obstacle recognition, lane marking identification, and drivable areas recognition based on a Deep Learning based Adaptive Termite Life Cycle Optimizer driven Efficient Residual Network (ATLC-EResNet) perception framework is established. Data pre-processing involves dealing with missing values to preserve trajectories, use of Kalman Filter for removing noise, and application of Min-Max Normalization for spatial and dynamic properties. The Dijkstra Algorithm (DA) search technique is used to minimize redundant searches and decrease spatial complexity. Reduction in redundant nodes using the ATCO improves the original trajectory obtained through minimization of control points.

Results

Results of dynamic scenarios demonstrate that the proposed method significantly outperforms path length (1050), speed (10,5) collision (1,8) and energy consumption (82,1) in terms of planning efficiency, safety, and path rationality. Specifically, the optimized algorithm decreases the number of traversed nodes, entire turning angle, and complete path length, while maintaining real-time performance.

Conclusion

The results confirm that the proposed ATLC-EResNet provides a robust and scalable solution for intelligent driving path planning in complex environments.

Zitieren

APA

Qian, D., & Lu, Q. (2026). Research on Optimization of Intelligent Driving Path Planning Algorithm Based on Deep Learning. Data and Metadata, 5, 806. https://doi.org/10.56294/dm2026806

Vancouver

Qian D, Lu Q. Research on Optimization of Intelligent Driving Path Planning Algorithm Based on Deep Learning. Data and Metadata. 2026;5:806. doi:10.56294/dm2026806

BibTeX
@article{Qian2026,
  author = {Qian, Dachuang and Lu, Qingshou},
  title = {Research on Optimization of Intelligent Driving Path Planning Algorithm Based on Deep Learning},
  journal = {Data and Metadata},
  year = {2026},
  volume = {5},
  pages = {806},
  doi = {10.56294/dm2026806}
}
Kostenlos starten →