Answers are grounded in this paper and cite their sources.
Three‐dimensional unmanned aerial vehicle (UAV) path planning in cluttered terrain requires a planner to search a continuous waypoint space, whereas the actual safety of a candidate can be assessed only after decoding the waypoint vector into a flight path. The decoded path must satisfy clearance, no‐fly‐zone, altitude, segment‐length and turning constraints, which makes the practical fitness landscape nonconvex, nonsmooth and strongly affected by narrow feasible corridors. This paper therefore studies UAV trajectory generation through a finite‐budget derivative‐free planning framework. An improved Exponential–Trigonometric Optimisation algorithm (IETO) is proposed for this real‐coded UAV waypoint model. The method combines optimised Latin hypercube sampling for initial spatial coverage, adaptive exploration–exploitation control for search‐radius regulation and probabilistic local refinement for terminal path improvement. Candidate paths are ranked by a nonsmooth exterior‐penalty fitness and are finally accepted only after explicit decoded‐path feasibility checks. Benchmark tests and deterministic, uncertain and long‐range UAV scenarios show that IETO produces high‐quality feasible trajectories and achieves the lowest mean planning fitness in the reported main UAV comparison under the same evaluation budget.
No takes yet. Share an insight, caveat, or question.
Ji et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: