MSGSO: A Multi-Strategy Glider Snake Optimizer for Global Optimization and 3D UAV Path Planning

Information in the Glider Snake Optimizer (GSO) propagates through a leader–predecessor chain. This structure is simple, but it can lose diversity when adjacent agents converge to the same region. The proposed Multi-Strategy Glider Snake Optimizer (MSGSO) retains the original GSO update and subsequently applies nonlinear dynamic polynomial mutation (NDM), elite opposition-based learning (EOBL), and quadratic interpolation (QI). MSGSO was evaluated on CEC 2019 and CEC 2022 with 30 matched random seeds and 15,000 objective evaluations for every algorithm–problem pair. The experiments included seven alternative optimizers and all single, pairwise, and three-operator GSO variants. Across the 34 benchmark problems, MSGSO significantly outperformed GSO on 31 and showed no significant loss. It nevertheless ranked third in each external comparison: L-SHADE led CEC 2019 and CEC 2022 at (Formula presented.), while CMA-ES led CEC 2022 at (Formula presented.). The ablation attributed most of the gain to NDM; NDM–GSO led the GSO variants on CEC 2019, and NDM–QI–GSO led at both CEC 2022 dimensions. In the UAV study, MSGSO returned 29 feasible paths in 30 runs in Scenario 1 and feasible paths in every run in the other two scenarios. It led the feasibility-first ranking in Scenarios 1 and 2 and placed third in Scenario 3. Thus, MSGSO improves its parent algorithm, although the full three-operator sequence is not consistently the best configuration.