# An Improved Red-Billed Blue Magpie Optimization Algorithm for 3D UAV Path Planning in Complex Terrain

## TL;DR

This paper introduces an improved optimization algorithm for 3D drone path planning in complex environments, enhancing performance and efficiency.

## Contribution

The CTWRBMO algorithm introduces four enhancement strategies to improve global search and avoid local optima in 3D UAV path planning.

## Key findings

- CTWRBMO outperforms existing algorithms in optimization accuracy and convergence speed.
- Paths planned with CTWRBMO are shorter, safer, and smoother in complex terrain simulations.
- The algorithm effectively balances exploration and exploitation for practical UAV applications.

## Abstract

This paper presents the Circle-Mapping Transition and Weighted Red-Billed Blue Magpie Optimizer (CTWRBMO), designed to address significant challenges in 3D path planning for drones. Although the original Red-Billed Blue Magpie Optimizer (RBMO) algorithm features a simple structure, few parameters, and strong local search capability, making it well-suited for UAV path optimization, it suffers from insufficient population diversity, limited global search ability, and a tendency to fall into local optima in complex high-dimensional scenarios. To overcome these limitations, four enhancement strategies are introduced. Firstly, the Circle chaotic mapping strategy leverages the randomness and ergodicity of chaotic sequences to generate an initial population that is uniformly distributed. This enhancement improves population diversity from the beginning and provides a solid foundation for global optimization. Secondly, the ε parameter is dynamically adjusted to prioritize local refinement during the early stages of optimization. This adjustment enables rapid convergence toward potentially optimal areas. This parameter increases to enhance global search capabilities as the algorithm progresses, thereby broadening the optimization space and achieving a dynamic equilibrium. Additionally, a nonlinear dynamic weighting factor (wd) is incorporated into the position update formula. The algorithm’s ability to escape local optima is significantly improved by dynamically altering the weight ratio between historical optimal positions and the current position. Furthermore, an elite perturbation mechanism based on individual neighborhoods is implemented to generate candidate solutions using local information. This mechanism enhances the algorithm’s local exploration capabilities and improves the stability of preserving optimal solutions, supported by a greedy criterion for optimal retention. Experimental results show that the CTWRBMO algorithm significantly outperforms comparison algorithms in terms of optimization accuracy and convergence speed, demonstrating exceptional global optimization capabilities. Additional applications in UAV 3D path planning simulations evaluated paths based on length, threat avoidance efficiency, and smoothness. The results indicate that paths planned using CTWRBMO are shorter, safer, and smoother compared to those generated by the Harrier Hawks Optimization (HHO), African Vulture Optimization Algorithm (AVOA), Artificial Bee Colony (ABC) Algorithm, and the traditional Magpie Algorithm, effectively meeting practical engineering requirements for UAV 3D path planning.