Work

Intelligent multi-UAV path planning in complex environments: a hybrid optimization approach with direction-assisted search and adaptive refinement

Year: 2025

Type: article

Abstract: Multi-UAV cooperative path planning in complex urban environments faces critical challenges in balancing exploration efficiency, convergence precision, and computational practicality. This study proposes a Hybrid Direction-Assisted Polar Light Optimization (HDPLO) algorithm that integrates three key innovations: directional heuristic mechanisms for intelligent search space expansion, adaptive crossover strategies for diversity maintenance, and CMA-ES integration for high-precision local refinement. Experimental validation using CEC2017 benchmarks and realistic multi-UAV scenarios (obstacle densities 7.71%-77.11%) demonstrates significant improvements: 38.9% better solution quality on benchmark functions, 45.7% faster convergence, 15.2% reduced path costs in UAV planning, and 11.1%-38.5% fewer turning maneuvers compared to state-of-the-art algorithms. Statistical analysis confirms performance reliability (p<0.001), while ablation studies verify each component’s essential contribution. These findings establish HDPLO as a robust solution for intelligent UAV systems requiring high reliability and efficiency in real-world deployments.

Source: Journal of King Saud University - Computer and Information Sciences

DOI: 10.1007/s44443-025-00398-4

Authors: Tiancheng Jin, Yujie Zhu

Institutions: Tongji University, East China Normal University

Language: English

Topic: Metaheuristic Optimization Algorithms Research

Subfield: Artificial Intelligence

Field: Computer Science

Domain: Physical Sciences

SDGs: Sustainable cities and communities, Industry, innovation and infrastructure

Open Access status: gold

Estimated APC paid: $1,990