Executable 3D Path Planning with Constrained Lazy Theta Star

This paper presents a constrained three-dimensional path planning method for low-altitude unmanned aerial vehicles operating in complex mountainous terrain environments. The proposed approach aims to generate collision-free and kinematically executable paths while maintaining real-time performance under strict motion constraints by explicitly incorporating vehicle pitch-angle and turn-angle limitations, directional guidance constraints, and safety distance requirements into the planning process. Motion feasibility is ensured during the search stage rather than through post-processing or trajectory correction, enabling reliable path generation in dense terrain with limited maneuverability and complex three-dimensional obstacle distributions. To this end, an integrated planning framework based on a constrained Lazy Theta* algorithm is developed, in which motion constraints are embedded into both node expansion and line-of-sight evaluation, together with a dual-resolution planning strategy that balances global search efficiency and local path feasibility. Extensive simulations in representative mountainous environments demonstrate that the proposed approach can robustly generate feasible three-dimensional trajectories while achieving real-time planning performance suitable for repeated planning in practical deployment scenarios. Furthermore, the framework exhibits adaptive constraint-aware search behavior, effectively balancing feasibility enforcement and computational efficiency across varying terrain densities and maneuverability limits.