Enhanced Secretary Bird Algorithm Improves Three-Dimensional Flight Path Planning in Simulations

A research team has introduced an adaptive metaheuristic framework designed to calculate safe, efficient three-dimensional trajectories for unmanned aerial vehicles operating in complex environments. Described in a regular research paper published in the International Journal of Intelligent Robotics and Applications, the method addresses spatial navigation challenges tied to the growth of the low-altitude economy. The proposed system, designated AGMSBOA-3D, extends the bio-inspired Secretary Bird Optimization Algorithm into volumetric space while integrating multi-strategy evolutionary enhancements to prevent collisions and streamline flight paths.

To transition the baseline optimization scheme from two dimensions into spatial navigation, the researchers formulated volumetric obstacle modeling, elevation-aware position update mechanisms, and three-dimensional collision detection protocols.

A central methodological component of the architecture is a composite objective function constructed to balance traversal efficiency against physical safety. This formulation simultaneously targets path length reduction and obstacle clearance by incorporating an adaptive reciprocal penalty term.

Beyond dimensional expansion and objective modeling, the framework integrates three distinct algorithmic enhancements derived from established metaheuristic design principles. These additions comprise adaptive parameter control, teaching-learning-based knowledge transfer, and an island-model multi-population evolutionary architecture. The authors emphasize that the principal novelty does not stem from these individual optimization concepts in isolation, but rather from their formal mathematical coupling and empirical performance gains within a three-dimensional path-planning formulation.

Within this coupled design, an adaptive learning strategy actively regulates step size throughout the search by monitoring population diversity and overall optimization progress. In parallel, a group learning strategy organizes structured knowledge transfer between identified elite agents and ordinary individuals across the population. Furthermore, the multi-population evolutionary strategy maintains specialized subpopulations that run concurrently, allowing the algorithm to balance broad exploration of unknown airspace with intensive local exploitation of known viable corridors.

The efficacy and necessity of these coupled mechanisms were examined through ablation testing and simulated trials across six distinct spatial environments characterized by varying obstacle densities. Across all benchmark comparisons, AGMSBOA-3D produced the highest safety margin and the smoothest flight trajectories among the evaluated algorithms. The experimental trials also demonstrated competitive convergence speed and computational runtime, alongside low run-to-run variability under the adopted protocol, though the authors note the algorithm did not claim the uniquely lowest standard deviation for every recorded metric.

The authors delineate clear boundaries regarding the current experimental scope and operational assumptions of their navigation model. All reported trajectories and comparative metrics were established under static, fully known obstacle configurations, excluding unanticipated hazards or moving objects. Furthermore, the present formulation does not yet integrate unmanned aerial vehicle kinematic constraints, such as specific turning radiuses, climb rates, or aerodynamic forces. The research team identified physical hardware validation and adaptation to dynamic operational environments as essential directions for future investigation.

Subject of Research: Three-dimensional UAV obstacle-avoidance path planning using an enhanced secretary bird optimization algorithm

Article Title: AGMSBOA-3D: an adaptive group multi-strategy enhanced secretary bird optimization algorithm for three-dimensional obstacle-avoidance path planning

Article References: Jiang, Z., Yuan, C., Liu, Z., Liu, H., Li, X., & Tian, Z. (2026). AGMSBOA-3D: an adaptive group multi-strategy enhanced secretary bird optimization algorithm for three-dimensional obstacle-avoidance path planning. International Journal of Intelligent Robotics and Applications. https://doi.org/10.1007/s41315-026-00609-9

Image Credits: AI Generated

DOI: 10.1007/s41315-026-00609-9

Keywords: path planning, unmanned aerial vehicles, secretary bird optimization algorithm, obstacle avoidance, metaheuristic algorithms, three-dimensional simulation, low-altitude economy, spatial navigation, volumetric modeling, trajectory smoothing, adaptive parameter control, group learning, multi-population evolution, collision detection, objective function optimization, flight simulation